A

PROJECT REPORT

ON

“Doctor Receipt Analyzer”

SUBMITTED TO THE SAVITRIBAI PHULE PUNE UNIVERSITY
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE AWARD OF THE

BACHELOR OF COMPUTER ENGINEERING

BY

1. Avhad Roshan Prakash (SEAT NO- B190274204)

2. Patil Sudarshan Bhaskar (SEAT NO - B190274241)
3. Patait Sumit Bhausaheb (SEAT NO - B190274301)
4. Patil Kalpesh Avinash (SEAT NO - B190274238)

UNDER THE GUIDANCE OF

Prof. Prabha B. Gadakh

DEPARTMENT OF COMPUTER ENGINEERING

Sir Visvesvaraya Institute Of Technology, Nashik
A/p.Chincholi,Tal.Sinnar,Dist.Nashik - 422102(MS)India
YEAR 2023-24
 CERTIFICATE
This is to certify that the Project Entitled

“Doctor Receipt Analyzer ”

Submitted By

1. Avhad Roshan Prakash (SEAT NO - B190274204)

2. Patil Sudarshan Bhaskar (SEAT NO - B190274241)
3. Patait Sumit Bhausaheb (SEAT NO- B190274301)
4. Patil Kalpesh Avinash (SEAT NO - B190274238)

is a bonafide work carried out by them under the supervision of Prof. Prabha B Gadakh
and it is approved for the partial fulfillment of the requirement of Savtribai Phule Pune
university, Pune for the award of the degree of Bachelor of Engineering (Computer Engi-
neering)

Prof. Prabha B. Gadakh Prof. Sharad M. Rokade

Guide HOD
Dept. of Computer Engg. Dept. of Computer Engg.

Dr G. B Shinde
Principal
Sir Visvesvaraya Institute Of Technology, Nashik
 Savitribai Phule Pune University

CERTIFICATE

This is to Certify that

1. Avhad Roshan Prakash (SEAT NO - B190274204)

2. Patil Sudarshan Bhaskar (SEAT NO - B190274241)
3. Patait Sumit Bhausaheb (SEAT NO - B190274301)
4. Patil Kalpesh Avinash (SEAT NO - B190274238)

Student of B.E. Computer

was examined in the project Report entitled

“Doctor Receipt Analyzer ”

on . . ./. . . /2023

At
DEPARTMENT OF COMPUTER ENGINEERING,
SIR VISVESVARAYA INSTITUTE OF TECHNOLOGY, NASHIK
YEAR 2023-24

..................... .....................
Internal Examiner External Examiner
 Certificate By Guide

This is to certify that

1. Avhad Roshan Prakash (SEAT NO - B190274204)

2. Patil Sudarshan Bhaskar (SEAT NO - B190274241)
3. Patait Sumit Bhausaheb (SEAT NO - B190274301)
4. Patil Kalpesh Avinash (SEAT NO - B190274238)

Has completed the work under my guidance and that, I have verified the work for its
originality in documentation, problem statement, literature survey and conclusion presented
in work.

Place: SVIT, Nashik (Prof. Prabha B. Gadakh)

Date:
 ACKNOWLEDGEMENT

We take this opportunity to express our hearty thanks to all those who helped us for this
project. We express our deep sense of gratitude to our Project Guide Prof. Prabha B
Gadakh for her guidance and continuous motivation. We gratefully acknowledge the help
provided by him for improving of this project with great interest. We express our heartiest
thanks for those who encouraged us and give their innovative ideas and suggestions for this
project. Lastly, we would like to thanks to all teachers and all my friends who helped us
for this Project.

1.Avhad Roshan Prakash

2.Patil Sudarshan Bhaskar
3.Patait Sumit Bhausaheb
4.Patil Kalpesh Avinash

i
 ABSTRACT

In this innovative application, we present a Machine Learning-based Doctor Receipt Ana-

lyzer system empowered by Optical Character Recognition (OCR) technology. The system
represents a transformative leap in the processing of medical receipts by seamlessly integrat-
ing advanced OCR algorithms. Capable of extracting critical information from scanned or
photographed receipts, the solution automates the traditionally manual task of deciphering
handwritten or printed text, enhancing both accuracy and efficiency. Tailored for effortless
integration into healthcare administration, the system intelligently decodes essential details
such as patient names, dates, services provided, and associated costs. The synergy of ma-
chine learning and OCR not only streamlines billing processes for healthcare professionals
but also promises to elevate the overall efficiency of medical documentation and financial
management, marking a significant advancement in healthcare technology.
Keywords: OCR, receipt analyser,Machine Learning, NLP.

ii
 SYNOPSIS

TITLE: Doctor Receipt Analyzer.

DOMAIN: Machine Learning, NLP.
OBJECTIVES:

1. Automate Data Extraction: Eliminate manual efforts involved in extracting informa-

tion from doctor receipts by employing OCR to recognize and extract text seamlessly.

2. Enhance Accuracy: Leverage machine learning algorithms to improve the accuracy

of information extraction, especially in cases of handwritten or complex text formats,
reducing errors in data interpretation.

3. Streamline Workflow: Simplify and expedite the billing and record-keeping processes
for healthcare professionals by providing a seamless and automated system for man-
aging receipt data.

4. Intelligent Information Retrieval: Enable the system to intelligently identify and ex-
tract key details such as patient names, dates, services provided, and associated costs,
contributing to efficient healthcare administration.

5. Improve Financial Management: Contribute to more effective financial management

within healthcare institutions by providing a reliable and automated tool for process-
ing and organizing receipt data.

6. Adaptability: Design the system to be adaptable to various receipt formats, ac-

commodating the diversity in documentation styles used across different healthcare
providers.

7. Reduce Workload: Alleviate the burden on healthcare professionals by automating

the time-consuming task of manual data entry, allowing them to focus more on patient
care and other critical aspects of their roles.

8. Technological Advancement: Introduce a technologically advanced solution that har-

nesses the capabilities of OCR and machine learning to bring about a paradigm shift

iii
 in the way medical receipts are handled, aligning with the evolving landscape of
healthcare technology.

0.1 ABSTRACT

In this innovative application, we present a Machine Learning-based Doctor Receipt Ana-

lyzer system empowered by Optical Character Recognition (OCR) technology. The system
represents a transformative leap in the processing of medical receipts by seamlessly integrat-
ing advanced OCR algorithms. Capable of extracting critical information from scanned or
photographed receipts, the solution automates the traditionally manual task of deciphering
handwritten or printed text, enhancing both accuracy and efficiency. Tailored for effortless
integration into healthcare administration, the system intelligently decodes essential details
such as patient names, dates, services provided, and associated costs. The synergy of ma-
chine learning and OCR not only streamlines billing processes for healthcare professionals
but also promises to elevate the overall efficiency of medical documentation and financial
management, marking a significant advancement in healthcare technology.
Keywords: OCR, receipt analyser,Machine Learning, NLP.

iv
 INDEX

ACKNOWLEDGEMENT i

ABSTRACT ii

ABSTRACT iii
0.1 ABSTRACT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iv

1 INTRODUCTION 1
1.1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 AIM/OBJECTIVES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.3 MOTIVATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

2 LITERATURE SURVEY 4

3 PROBLEM STATEMENT 7
3.1 PROBLEM STATEMENT . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

4 SOFTWARE REQUIREMENT SPECIFICATION 8

4.1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
4.1.1 PROJECT SCOPE . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
4.1.2 USER CLASSES AND CHARACTERISTICS . . . . . . . . . . . . 10
4.1.3 ASSUMPTIONS AND DEPENDENCIES . . . . . . . . . . . . . . . 10
4.2 FUNCTIONAL REQUIREMENTS . . . . . . . . . . . . . . . . . . . . . . . 12
4.2.1 User Authentication and Authorization: . . . . . . . . . . . . . . . 12
4.2.2 Receipt Upload and Processing: . . . . . . . . . . . . . . . . . . . . . 12
4.2.3 OCR Integration: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
4.2.4 Data Extraction and Parsing: . . . . . . . . . . . . . . . . . . . . . . 12
4.2.5 Database Management: . . . . . . . . . . . . . . . . . . . . . . . . . 12
4.2.6 SYSTEM INTERFERENCE . . . . . . . . . . . . . . . . . . . . . . 12
4.3 EXTERNAL INTERFACE REQUIREMENTS . . . . . . . . . . . . . . . . 13

v
 4.3.1 USER INTERFACE . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
4.3.2 HARDWARE INTERFACE . . . . . . . . . . . . . . . . . . . . . . . 13
4.3.3 SOFTWARE INTERFACE . . . . . . . . . . . . . . . . . . . . . . . 13
4.3.4 COMMUNICATION INTERFACE . . . . . . . . . . . . . . . . . . . 13
4.4 USER INTERFACES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
4.4.1 SIX USABILITY GOALS . . . . . . . . . . . . . . . . . . . . . . . . 13
4.5 NONFUNCTIONAL REQUIREMENTS . . . . . . . . . . . . . . . . . . . . 14
4.5.1 Usability: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
4.5.2 Reliability: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
4.5.3 Performance: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
4.5.4 Scalability: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
4.5.5 Open standard: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
4.6 SYSTEM REQUIREMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . 15
4.7 Programming Language . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
4.7.1 Framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
4.7.2 Development Tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
4.7.3 Database . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
4.8 Hardware Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
4.9 Software Requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
4.10 ANALYSIS MODELS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

5 METHODOLOGY AND SYSTEM DESIGN 21

5.1 SYSTEM METHODOLOGY . . . . . . . . . . . . . . . . . . . . . . . . . . 21
5.1.1 User Interaction and Receipt Upload: . . . . . . . . . . . . . . . . . 21
5.1.2 Image Preprocessing: . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
5.1.3 OCR Processing: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
5.1.4 Text Extraction and Parsing: . . . . . . . . . . . . . . . . . . . . . . 22
5.1.5 Data Storage: . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
5.1.6 User Reporting and Analytics: . . . . . . . . . . . . . . . . . . . . . 22

6 PROJECT DESIGN 23
6.1 Data Flow Diagrams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
6.2 UML Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
6.2.1 Activity Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27
6.2.2 Use Case Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
6.2.3 Sequence Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30

vi
 6.2.4 Class Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
6.3 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32

7 METHODOLOGY AND SYSTEM DESIGN 33

7.1 SYSTEM ARCHITECTURE . . . . . . . . . . . . . . . . . . . . . . . . . . 33
7.2 SYSTEM MODULES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
7.2.1 User Interface Module: . . . . . . . . . . . . . . . . . . . . . . . . . . 35
7.2.2 Image Preprocessing Module: . . . . . . . . . . . . . . . . . . . . . . 35
7.2.3 OCR Processing Module: . . . . . . . . . . . . . . . . . . . . . . . . 35

8 PROJECT IMPLEMENTATION 36
8.1 OVERVIEW OF PROJECT MODULES . . . . . . . . . . . . . . . . . . . . 36
8.2 TOOLS AND TECHNOLOGY USED . . . . . . . . . . . . . . . . . . . . . 37
8.2.1 OCR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
8.3 System Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

9 DEPLOYMENT AND MAINTENANCE 40

9.1 Deployment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
9.2 Deployment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40

10 PROJECT ESTIMATION,SCHEDULE AND TEAM STRUCTURE 42

10.1 Project Estimation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
10.2 Reconciled Estimates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
10.2.1 Efforts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
10.2.2 Development time per month . . . . . . . . . . . . . . . . . . . . . . 43
10.2.3 Development time for Project . . . . . . . . . . . . . . . . . . . . . . 43
10.2.4 Number of Persons . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
10.2.5 Project Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44
10.3 Risk Management w.r.t. NP Hard analysis . . . . . . . . . . . . . . . . . . . 45
10.3.1 Risk Identification . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
10.3.2 Product Size Risk . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
10.3.3 Business Impact Rule . . . . . . . . . . . . . . . . . . . . . . . . . . 46
10.3.4 Process Risk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
10.3.5 Technical Risk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
10.3.6 Development Environment Risk . . . . . . . . . . . . . . . . . . . . . 47
10.3.7 Staff Size And Experience Risk . . . . . . . . . . . . . . . . . . . . . 47
10.3.8 Risk Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

vii
 10.4 Feasibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
10.4.1 Technical Feasibility . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
10.4.2 Cost Feasibility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
10.5 Project Schedule . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
10.5.1 Project task set . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
10.6 Project Plan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50

11 SOFTWARE TESTING AND VALIDATION 52

11.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
11.2 Types of Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52
11.2.1 Manual Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
11.2.2 Automated Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
11.2.3 Unit Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
11.2.4 Integration Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
11.2.5 Regression Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
11.3 Software testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
11.4 Black Box Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
11.4.1 Black-box . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
11.5 White Box Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
11.5.1 White-box . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57

12 RESULT AND ANALYSIS 59

12.1 System Layout . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59

13 ADVANTAGES LIMITATIONS AND APPLICATIONS 65

CONCLUSION 67

14 SUMMARY AND CONCLUSION 68

REFERENCES 68

Appendix 70
A DETAILS OF THE PAPERS PUBLICATION 70

B PLGARISIM REPORT 91

viii
 List of Figures

4.1 VS Code programming language . . . . . . . . . . . . . . . . . . . . . . . . 15

4.2 Python software programming language . . . . . . . . . . . . . . . . . . . . 18
4.3 SQLite Database . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

6.1 DFD 0 Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

6.2 DFD 1 Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
6.3 DFD 2 Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
6.4 Activity Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
6.5 Usecase Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
6.6 Sequence Diagram . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

7.1 System Architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

10.1 NP-hard and NP-complete . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45

ix
 List of Tables

10.1 Hardware Resources Required . . . . . . . . . . . . . . . . . . . . . . . . . . 44

10.2 Risk Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
10.3 Probability and Value . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
10.4 Impact and Value . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
10.5 Planner and Progress Report for project . . . . . . . . . . . . . . . . . . . . 51

x
 Chapter 1

INTRODUCTION

1.1 INTRODUCTION

Introducing a cutting-edge solution, our Machine Learning-based Doctor Receipt Analyzer

system harnesses the power of OCR technology to revolutionize the way medical receipts
are processed. By leveraging sophisticated Optical Character Recognition algorithms, the
system adeptly extracts crucial information from scanned or photographed receipts. This
innovative approach not only automates the traditionally manual and time-consuming task
of deciphering handwritten or printed text but also enhances accuracy and efficiency. Seam-
lessly integrating into healthcare administration, the system intelligently decodes patient
names, dates, services provided, and associated costs. The synergy of machine learning and
OCR not only streamlines the billing and record-keeping process for healthcare professionals
but also promises to elevate the overall efficiency of medical documentation and financial
management.
This groundbreaking Machine Learning-driven Doctor Receipt Analyzer signifies a paradigm
shift in healthcare management. By seamlessly merging the capabilities of OCR technology
with advanced algorithms, our system revolutionizes the processing of medical receipts. Its
prowess lies in its ability to adeptly extract vital information from diverse receipt formats,
be they scanned or photographed. Beyond automating the traditionally labor-intensive
task of deciphering handwritten or printed text, the system significantly boosts accuracy
and operational efficiency. Tailored to seamlessly integrate into healthcare administration
workflows, it intelligently decodes key elements such as patient names, dates, services ren-
dered, and associated costs. This harmonious blend of machine learning and OCR not only
simplifies billing and record-keeping for healthcare professionals but also holds the promise
of elevating the overall efficiency of medical documentation and financial management to
unprecedented levels.

1
 Doctor Prescription Analyzer

1.2 AIM/OBJECTIVES

1. Automate Data Extraction: Eliminate manual efforts involved in extracting informa-

tion from doctor receipts by employing OCR to recognize and extract text seamlessly.

2. Enhance Accuracy: Leverage machine learning algorithms to improve the accuracy

of information extraction, especially in cases of handwritten or complex text formats,
reducing errors in data interpretation.

3. Streamline Workflow: Simplify and expedite the billing and record-keeping processes
for healthcare professionals by providing a seamless and automated system for man-
aging receipt data.

4. Intelligent Information Retrieval: Enable the system to intelligently identify and ex-
tract key details such as patient names, dates, services provided, and associated costs,
contributing to efficient healthcare administration.

5. Improve Financial Management: Contribute to more effective financial management

within healthcare institutions by providing a reliable and automated tool for process-
ing and organizing receipt data.

6. Adaptability: Design the system to be adaptable to various receipt formats, ac-

commodating the diversity in documentation styles used across different healthcare
providers.

7. Reduce Workload: Alleviate the burden on healthcare professionals by automating

the time-consuming task of manual data entry, allowing them to focus more on patient
care and other critical aspects of their roles.

8. Technological Advancement: Introduce a technologically advanced solution that har-

nesses the capabilities of OCR and machine learning to bring about a paradigm shift
in the way medical receipts are handled, aligning with the evolving landscape of
healthcare technology.

1.3 MOTIVATION

The motivation behind developing a Machine Learning-based Doctor Receipt Analyzer sys-
tem using OCR technology stems from the desire to address significant challenges in the
current healthcare administration landscape. Manual processing of doctor receipts is time-
consuming and prone to errors. The motivation is to enhance the efficiency of healthcare

2
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Doctor Prescription Analyzer

administration by automating the extraction of crucial information from receipts, reduc-

ing the burden on healthcare professionals. Handwritten or varied receipt formats can
pose challenges in accurate data extraction. By integrating OCR technology and machine
learning algorithms, the system aims to improve accuracy in interpreting and extracting
information, minimizing errors in critical details. The motivation is to create a solution
that seamlessly integrates into existing healthcare workflows, providing a user-friendly and
efficient tool for processing receipts. This streamlining contributes to more effective and
organized healthcare management. Automating the receipt analysis process can lead to
cost savings by reducing the need for manual data entry and potential errors. This moti-
vation aligns with the broader goal of optimizing financial management within healthcare
institutions. The motivation is to leverage cutting-edge technologies such as OCR and ma-
chine learning to bring about a technological advancement in healthcare administration.
This system represents an innovative approach to handling and analyzing doctor receipts,
aligning with the evolving landscape of healthcare technology.

3
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Chapter 2

LITERATURE SURVEY

This chapter discuss brief literature regarding the project. Literature survey is mainly used
to identify information relevant to the project work and know impact of it within the project
area. It defines as till yet how many surveys have been done knowledge of latest technology
and implementation designs.

PAPER 1

1. Yongjian Ren; Yuliang Shi, A Drug Recommendation Model Based on Message Propa-
gation and DDI Gating Mechanism, IEEE Journal of Biomedical and Health Informatics (
Volume: 26, Issue: 7, July 2022): Drug recommendation task based on the deep learning
model has been widely studied and applied in the health care field in recent years. However,
the accuracy of drug recommendation models still needs to be improved. In addition, the
existing recommendation models either give only one recommendation (however, there may
be a variety of drug combination options in practice) or can not provide the confidence
level of the recommended result. To fill these gaps, a Drug Recommendation model based
on Message Propagation neural network (denoted as DRMP) is proposed in this paper.
Then, the Drug-Drug Interaction (DDI) knowledge is introduced into the proposed model
to reduce the DDI rate in recommended drugs. Finally, the proposed model is extended to
Bayesian Neural Network (BNN) to realize multiple recommendations and give the confi-
dence of each recommendation result, so as to provide richer information to help doctors
make decisions. Experimental results on public data sets show that the proposed model is
superior to the best existing models.[1].

PAPER 2

2. Rika Fitriani; Faiza Renaldi; Irma Santikarama; Wu Zhou, Recommendation On Drugs

4
 Doctor Prescription Analyzer

Prescription For Diabetic Patient Using Case based Reasoning, 2022 International Confer-
ence on Science and Technology (ICOSTECH): As medical workers, we are often required
to be maximal in treating patients. The knowledge possessed by medical personnel, in this
case, doctors, must be demanded to the maximum so that there are no errors in providing
treatment to patients. Therefore we need a system to minimize the mistakes, namely a rec-
ommendation system. So this study builds a drug recommendation system for patients who
have diabetes. One method applied in a recommendation is Case Base Reasoning (CBR).
The CBR method calculates based on the similarity of the old case to the new issue. In
calculating the similarity of the problems using the Nearest Neighbors algorithm. Although
there have been many studies on drug recommendations, no studies have added risk factor
variables. Risk factors can influence results more accurately. The results of this study are
in the form of a prescription drug recommendation system for diabetic patients based on
the type of diabetes they suffer from. They are testing the accuracy of this system using
the Confusion Matrix with an accuracy of 80.60%. Suggestions for further research are to
add several types of similarity. It is necessary to research different kinds of diseases. It
is expected to produce an integrated system that can provide drug recommendations for
various conditions.[2].

PAPER 3

3. Md. Deloar Hossain; Md. Shafiul Azam, Drugs Rating Generation and Recommen-
dation from Sentiment Analysis of Drug Reviews using Machine Learning, 2020 Emerging
Technology in Computing, Communication and Electronics (ETCCE): A recommendation
system can assist the user to compose an understanding of requirements and propose in-
formed decisions from a lot of complicated knowledge. Recommendation from an analysis
of sentiments seems to be a great challenge as user-generated content is represented using
human language in several complicated ways. Many studies have focused on common fields
such as reviews of electrical items, films, and restaurants, but not enough on health and
medical issues. Sentiment analysis of healthcare in general and that of the drug experiences
of individuals, in particular, may shed considerable light on how to focus on improving
public health and reach the correct decision. In this paper, we design and implement a
drug recommender system framework that applies sentiment analysis technologies on drug
reviews. The objective of this research is to build a decision-making support platform to
help patients to achieve more significant choices in drug selection. Firstly, we propose a sen-
timental measurement approach to drug reviews and generate ratings on drugs. Secondly,
we take how much the drug reviews are useful to users, patient’s conditions, and dictionary

5
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Doctor Prescription Analyzer

sentiment polarity of drug reviews into consideration. Then, we fuse those factors into the
recommendation system to list appropriate medications. Experiments have been carried
out using Decision Tree, K-Nearest Neighbors, and Linear Support Vector Classifier algo-
rithm in rating generation and Hybrid model in recommendation based on the given open
dataset. The analysis is carried out to tune the parameters for each algorithm in order to
achieve greater performance. Finally, Linear Support Vector Classifier is selected for rating
generation to obtain a good trade-off among model accuracy, model efficiency, and model
scalability.[3].

PAPER 4

4. Tianqi Jia; Cuiying Wang, Knowledge Discovery of Prescriptions and Similar Pre-
scriptions of Guizhi Decoction Based on Structural Partial-Ordered Attribute Diagram,
2013 Third International Conference on Instrumentation, Measurement, Computer, Com-
munication and Control: As a new tool of data analysis and knowledge processing, formal
concept analysis has drawn more and more attention in various fields. This research fo-
cuses on prescriptions of GuiZhi Decoction, which is from Zhongjing Zhang’s Treatise on
Cold Pathogenic Diseases. First, we constructed the database for prescriptions of GuiZhi
Decoction based on the prescriptions’ names, the concept extraction and formal expression.
Then we generate the formal context and optimized it by use of the mathematical theory of
formal concept analysis and the relationship between prescriptions and herbs in database.
In the formal context, the prescriptions are objects and the herbs are attributes. Next the
structural partial-ordered attribute diagram is generated, so as to visualize the knowledge
of prescriptions of GuiZhi Decoction. Finally, through the analysis of structural partial-
ordered attribute diagram, more comprehensive, objective and multileveled knowledge are
discovered.[4].

6
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Chapter 3

PROBLEM STATEMENT

3.1 PROBLEM STATEMENT

The current manual process of handling doctor receipts in the healthcare system presents
a pressing set of challenges that hinder operational efficiency and accuracy. The labor-
intensive nature of manual data entry results in errors and delays, particularly when deci-
phering handwritten or varied receipt formats. This inefficiency not only impacts the timely
billing processes but also places a significant burden on healthcare professionals, diverting
their attention from critical patient care responsibilities. The lack of standardized receipt
formats across healthcare providers adds complexity, leading to inconsistent data interpre-
tation. Moreover, the absence of advanced technological solutions in the current landscape
overlooks the potential benefits of automation, machine learning, and Optical Character
Recognition (OCR) technology. These challenges collectively contribute to financial dis-
crepancies, jeopardizing the overall financial management of healthcare institutions. The
Doctor Receipt Analyzer system aims to address these issues by introducing an automated
solution that utilizes machine learning and OCR technology to streamline the processing of
doctor receipts, mitigate errors, and enhance the efficiency of healthcare administration.

7
 Chapter 4

SOFTWARE REQUIREMENT SPECIFICATION

This chapter is about the software requirement specification.It contain the overall prob-
lem Statement of the project and the hardware requirements, software requirements and
functional and non-functional requirements of the project.

4.1 INTRODUCTION

A Software Requirements Specification (SRS) is a document that describes the nature of a

project, software or application. In simple words, SRS document is a manual of a project
provided it is prepared before you kick-start a project/application. This document is also
known by the names SRS report, software document. A software document is primarily
prepared for a project, software or any kind of application.
The Doctor Receipt Analyzer system will be developed with the following software re-
quirements to ensure a robust and effective solution. The system will employ Optical
Character Recognition (OCR) technology and machine learning algorithms, and therefore,
it necessitates compatibility with OCR libraries such as Tesseract. The application will
be designed to run on major operating systems, including Windows, macOS, and Linux,
ensuring widespread accessibility. A web-based user interface will be developed, requir-
ing compatibility with modern web browsers such as Google Chrome, Mozilla Firefox, and
Safari. The system will store and manage data, demanding integration with a relational
database management system (RDBMS) like MySQL or PostgreSQL. Additionally, the ap-
plication will be built using programming languages such as Python and frameworks like
Django or Flask for the backend, and HTML, CSS, and JavaScript for the frontend. To fa-
cilitate version control and collaborative development, a Git repository hosted on platforms
like GitHub will be utilized. Regular updates and maintenance will be managed through a
continuous integration/continuous deployment (CI/CD) pipeline. These software require-
ments collectively form the foundation for a scalable, efficient, and user-friendly Doctor

8
 Doctor Prescription Analyzer

Receipt Analyzer system.

4.1.1 PROJECT SCOPE

The scope of the Machine Learning-based Doctor Receipt Analyzer system using OCR tech-
nology is expansive and transformative within the realm of healthcare administration. By
automating the extraction of crucial information from doctor receipts, the system aims to
revolutionize the traditionally labor-intensive and error-prone process of manual data en-
try. Its adaptability to diverse receipt formats, including handwritten and printed text,
ensures a broad applicability across different healthcare providers. The system’s potential
to enhance accuracy, particularly in deciphering complex formats, addresses a persistent
challenge in data interpretation. Beyond mere automation, its integration into existing
healthcare workflows streamlines processes, expediting billing procedures and reducing ad-
ministrative burdens on healthcare professionals. The system’s scope extends to contribut-
ing to more effective financial management within healthcare institutions, introducing ad-
vanced technologies like OCR and machine learning to advance the industry’s technological
landscape. Moreover, by freeing up valuable time for healthcare professionals, the system
fosters an environment conducive to improved patient care, aligning with the broader goal
of enhancing overall healthcare efficiency and quality.

9
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Doctor Prescription Analyzer

4.1.2 USER CLASSES AND CHARACTERISTICS

1. Healthcare Professionals:-
Characteristics: Primarily doctors, nurses, and administrative staff responsible for manag-
ing patient records and billing.

2. Healthcare Administrators: -
Characteristics: Administrative personnel overseeing financial management, reporting, and
overall operational efficiency.

3. Patients:-
Characteristics: Individuals receiving medical services and generating receipts for health-
care transactions.

4. System Administrators:-
Characteristics: Professionals overseeing the overall functioning and configuration of the
Machine Learning-based Doctor Receipt Analyzer system.

4.1.3 ASSUMPTIONS AND DEPENDENCIES

Assumptions:

1. Receipt Legibility: It is assumed that the doctor receipts provided for analysis are suf-
ficiently legible, allowing OCR technology to accurately extract information. Hand-
written text should be reasonably clear to ensure optimal performance.

2. Standardized OCR Library: The system assumes the availability and successful inte-
gration of a standardized OCR library, such as Tesseract, for accurate text extraction.
The performance of the OCR technology is contingent on the capabilities and accu-
racy of the selected library.

3. Consistent Receipt Formats: The system assumes a certain level of consistency in

receipt formats across different healthcare providers. While designed to handle diverse
formats, extreme variations may impact the accuracy of information extraction.

4. Internet Connectivity: Assuming the system is designed as a web-based application,

it is expected that users have reliable internet connectivity for accessing and utilizing
the system.

10
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Doctor Prescription Analyzer

5. Data Privacy Compliance: The system assumes compliance with data privacy regula-
tions, such as HIPAA, to ensure the secure handling of sensitive patient information.
Proper encryption and security measures are implemented to protect the confiden-
tiality of healthcare data.

Dependencies:

1. OCR Library Integration: The successful integration of the chosen OCR library is a
critical dependency for the accurate extraction of information from doctor receipts.
Any changes or updates to the OCR library may impact the system’s performance.

2. Web Browser Compatibility: The system’s web-based user interface depends on com-
patibility with modern web browsers, and any changes in browser behavior may affect
the user experience.

3. Database Management System (DBMS): The system relies on a relational database

management system (e.g., MySQL or PostgreSQL) for data storage. Any changes to
the DBMS or database structure may have dependencies on the system’s functionality.

4. Technological Frameworks: The system’s development may depend on specific techno-

logical frameworks, such as Django or Flask for the backend, and HTML, CSS, and
JavaScript for the frontend. Updates or changes to these frameworks may impact
system compatibility.

5. Regulatory Compliance Updates: Dependencies on adherence to data privacy regula-

tions, such as HIPAA, require the system to stay abreast of any changes in compliance
requirements, impacting system features and security measures.

11
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Doctor Prescription Analyzer

4.2 FUNCTIONAL REQUIREMENTS

4.2.1 User Authentication and Authorization:

Implement a secure user authentication system to ensure authorized access for healthcare
professionals, administrators, and other users.

4.2.2 Receipt Upload and Processing:

Allow users to upload scanned or photographed receipts for processing.

4.2.3 OCR Integration:

Integrate a robust OCR library (e.g., Tesseract) to accurately extract text from receipt
images.

4.2.4 Data Extraction and Parsing:

Extract relevant information from the OCR-processed text, including patient names, dates,
services provided, and associated costs.

4.2.5 Database Management:

Store and manage extracted data in a relational database for retrieval and analysis.

4.2.6 SYSTEM INTERFERENCE

1 The AI-Powered Placement Management System should integrate smoothly with ex-
isting HR systems, such as human resource information systems (HRIS) and payroll
software.

2 Interference during data migration can lead to data loss or inaccuracies, affecting
historical records and the overall integrity of the system.

3 The quality and availability of training data can interfere with the performance of AI
models.

12
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Doctor Prescription Analyzer

4.3 EXTERNAL INTERFACE REQUIREMENTS

4.3.1 USER INTERFACE

User has to interface with web based application to access the features and to provide easy
communication with system.

4.3.2 HARDWARE INTERFACE

The computer system is used to control the system.

4.3.3 SOFTWARE INTERFACE

The web browser is used to interface with the web application.

4.3.4 COMMUNICATION INTERFACE

There is a specific network protocol as long as the performance requirement are satisfied.

4.4 USER INTERFACES

The user experience should be considered as priority in user interface. This is the way that
the product will be used by users. Users should meet the exact needs they want, without
confuse. Designer should clear the primary objective of developing an interactive product.
It is suggested to classify the objectives in terms of usability and user experience goals.
There are six goals of usability. They could make the product easy to learn and effective to
use.

4.4.1 SIX USABILITY GOALS

1. Effective to use(effectiveness)

2. Efficient to use(efficiency)

3. Safe to use(safety)

4. Having good utility(utility)

5. Easy to learn(learnability)

6. Easy to remember how to use(memorability)

13
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Doctor Prescription Analyzer

Effectiveness is a common goal to reach the best result of the expectation. The performance
of the software is satisfactory. Efficiency is focus on the cost of computation of the software.
Most users make an attention on the speed of software, they think every action should be
fluent. If a lag accrued during the operation, people will think there are some problems
with it. It will worsen the user experience.

4.5 NONFUNCTIONAL REQUIREMENTS

4.5.1 Usability:

The ease with which the system can be learned, managed or used. Usability gives the
measure of how much user friendly the system is.

4.5.2 Reliability:

The degree to which the system must work for users. It also refers to the mean time between
failures, means what can be the maximum down time.

4.5.3 Performance:

Performance specifications typically refer to response time, transaction throughput, and

capacity. They deal with response time, which means the time taken by the system to load,
reload, screen open and refresh times etc.

4.5.4 Scalability:

It refers to the ability of the proposed software application to increase the number of users
or applications associated with the product.

4.5.5 Open standard:

To ensures the viability and future expansion of the system, all offered development tools,
server software, as well as, the application are based on open templates and are available
under the terms of the General Public License.

14
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Doctor Prescription Analyzer

4.6 SYSTEM REQUIREMENTS

4.7 Programming Language

Visual Studio Code is a free and open-source code editor developed by Microsoft. It is a
popular choice among developers for its versatility and extensive set of features.

Figure 4.1: VS Code programming language

Here are some key aspects of Visual Studio Code (VS Code):

1. Cross-Platform: VS Code is available for Windows, macOS, and Linux, making it a

versatile choice for developers on various platforms.

2. Lightweight: It’s a lightweight code editor that’s faster and uses fewer system re-
sources compared to full-fledged integrated development environments (IDEs).

3. Extensible: VS Code supports a wide range of programming languages and has a rich
extension ecosystem. You can install extensions to tailor the editor to your specific
needs.

4. Intelligent Code Editing: It offers features like syntax highlighting, code completion,
and linting for various programming languages.

5. Integrated Git: VS Code has built-in Git support, making it easy to work with version
control for your projects.

6. Debugging: It provides integrated debugging tools for various languages and plat-
forms.

15
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Doctor Prescription Analyzer

7. Terminal Integration: You can run terminal commands directly within the editor.

8. Customizable Themes and Styles: VS Code allows you to choose from a variety of
themes and customize the editor’s appearance to your liking.

9. Community and Documentation: There’s a large and active community around VS

Code, and you can find extensive documentation and tutorials to help you get started.

10. Visual Studio Live Share: A feature that allows real-time collaboration between de-
velopers, making it easier to work on code together.

VS Code has gained popularity in a wide range of development communities, from web
development to data science, and it’s known for its ease of use, performance, and the ability
to adapt it to different coding workflows.

16
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Doctor Prescription Analyzer

4.7.1 Framework

Django is a high-level Python web framework that encourages rapid development and clean,
pragmatic design. It follows the Model-View-Controller (MVC) architectural pattern and
is designed to make it easier for developers to build web applications quickly and efficiently.
Django provides a wide range of features and tools, including:

1. An Object-Relational Mapping (ORM) system for database management.

2. A built-in admin interface for managing the application’s data.

3. A URL routing system.

4. A template system for creating dynamic web pages.

5. Authentication and authorization mechanisms.

6. Security features to protect against common web application vulnerabilities.

7. Internationalization and localization support.

8. Extensibility through reusable apps and plugins.

9. Community support and a vast ecosystem of third-party packages.

Django is commonly used for developing all sorts of web applications, from small personal
projects to large, complex websites. It is known for its robustness, scalability, and the
”batteries-included” philosophy, which means that it comes with a lot of built-in function-
ality to help you get started quickly.

17
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Doctor Prescription Analyzer

4.7.2 Development Tools

Figure 4.2: Python software programming language

Python is a multi-paradigm programming language. Object-oriented programming and

structured programming are fully supported, and many of their features support functional
programming and aspect-oriented programming (including metaprogramming and metaob-
jects).
Python is an interpreted, object-oriented, high-level programming language with dynamic
semantics. Its high-level built in data structures, combined with dynamic typing and dy-
namic binding, make it very attractive for Rapid Application Development, as well as for use
as a scripting or glue language to connect existing components together. Python’s simple,
easy to learn syntax emphasizes readability and therefore reduces the cost of program main-
tenance. Python supports modules and packages, which encourages program modularity
and code reuse.

18
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Doctor Prescription Analyzer

4.7.3 Database

Figure 4.3: SQLite Database

SQLite is an embedded, server-less relational database management system. It is an in-

memory open-source library with zero configuration and does not require any installation.
Also, it is very convenient as it’s less than 500kb in size, which is significantly lesser than
other database management systems.
Features of SQLite as follows,

• SQLite is used to develop embedded software for devices like televisions, cell phones,
cameras, etc.

• It Can manage low to medium-traffic HTTP requests.

• SQLite can change files into smaller size archives with lesser metadata.

• SQLite is used as a temporary dataset to get processed with some data within an
application.

• SQLite is an open-source software. The software does not require any license after
installation. SQLite is serverless as it doesn’t need a different server process or system
to operate.

• SQLite facilitates you to work on multiple databases on the same session simultane-
ously, thus making it flexible.

19
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Doctor Prescription Analyzer

4.8 Hardware Requirements

• AMD/Intel i3 Processor or above Processor

• 4GB RAM for application development

• 150 GB or above Hard Disk

4.9 Software Requirements

• Windows 7 or above

• Vscode, Xamp

• Python

• Django

4.10 ANALYSIS MODELS

Analysis model operate as a link between the system description and the design model.
In the analysis model information function and the behavior of the system is define and
these are translated into the architecture interface and component level design in the design
modeling.

20
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Chapter 5

METHODOLOGY AND SYSTEM DESIGN

5.1 SYSTEM METHODOLOGY

The system methodology for the Machine Learning-based Doctor Receipt Analyzer, utilizing
OCR technology, follows a systematic and iterative process to ensure accurate and efficient
extraction of information from medical receipts. The methodology can be broken down into
the following key steps:

5.1.1 User Interaction and Receipt Upload:

Users interact with the system through a user-friendly web interface. They upload scanned
or photographed receipts into the system for processing. Implement a secure and intuitive
user interface with an option for file upload. Validate user credentials and ensure proper
access control.

5.1.2 Image Preprocessing:

Receipt images undergo preprocessing to enhance OCR accuracy. This step involves resizing,
normalization, and contrast adjustments. Utilize image processing techniques to prepare
receipt images for optimal OCR performance. This ensures that the OCR engine receives
clear and standardized inputs.

5.1.3 OCR Processing:

Implement OCR technology (e.g., Tesseract) to convert the preprocessed receipt images
into machine-readable text. Integrate the chosen OCR library into the system, configure it
to handle various receipt formats, and initiate the process of text extraction.

21
 Doctor Prescription Analyzer

5.1.4 Text Extraction and Parsing:

Extract relevant information from the OCR-processed text, including patient names, dates,
services provided, and associated costs. Develop algorithms for parsing and pattern recog-
nition to identify key data fields within the extracted text. Implement error-checking mech-
anisms to ensure accuracy.

5.1.5 Data Storage:

Store the extracted data in a secure and structured manner within a relational database for
future retrieval and analysis. Integrate with a database management system (e.g., MySQL
or PostgreSQL) to create a schema for receipt information. Implement secure data storage
and retrieval mechanisms.

5.1.6 User Reporting and Analytics:

Provide users with tools to generate reports and gain insights from the processed receipt
data.

22
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Chapter 6

PROJECT DESIGN

6.1 Data Flow Diagrams

A data flow diagram (DFD) is a graphical or visual representation using a standardized

set of symbols and notations to describe a business’s operations through data movement.
They are often elements of a formal methodology such as Structured Systems Analysis and
Design Methods.
The objective of a DFD is to show the scope and boundaries of a system as a whole. It
may be used as a communication tool between a system analyst and any person who plays
a part in the order that acts as a starting point for redesigning a system. The DFD is also
called as a data flow graph or bubble chart.

23
 Doctor Prescription Analyzer

DFD 0, also called context diagram of the result management system. As the
bubbles are decomposed into less and less abstract bubbles, the corresponding data flow
may also be needed to be decomposed.

Figure 6.1: DFD 0 Diagram

24
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Doctor Prescription Analyzer

DFD 1, a context diagram is decomposed into multiple bubbles/processes. In this

level, we highlight the main objectives of the system and breakdown the high-level process
of 0-level DFD into subprocesses.

Figure 6.2: DFD 1 Diagram

25
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Doctor Prescription Analyzer

DFD 2 goes one process deeper into parts of 1-level DFD. It can be used to project
or record the specific/necessary detail about the system’s functioning.

Figure 6.3: DFD 2 Diagram

26
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Doctor Prescription Analyzer

6.2 UML Diagram

6.2.1 Activity Diagram

Use cases show what your system should do. Activity diagrams allow you to specify
how your system will accomplish its goals. Activity diagrams show high-level actions chained
together to represent a process occurring in your system. An activity diagram is essentially
a flowchart, showing flow of control from activity to activity. Unlike a traditional flowchart,
an activity diagram shows concurrency as well as branches of control. Activity diagrams
focus on the dynamic flow of a system.

27
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Doctor Prescription Analyzer

Figure 6.4: Activity Diagram

28
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Doctor Prescription Analyzer

6.2.2 Use Case Diagram

Four modeling elements make up the use case diagram; these are:

• Actors: Actors refer to a type of users, users are people who use the system. In this
case student, teacher developer are the users of the framework and application

• Use cases: A use case defines behavioral features of a system. Each use case is
named using a verb phrase that express a goal of the system. The name may appear
inside or outside the ellipse.

• Associations: An association is a relationship between an actor and a use case. The

relationship is represented by a line between an actor and a use case.

• The include relationship: It is analogous to a call between objects. One use case
requires some type of behavior which is fully defined in another use case.

29
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Doctor Prescription Analyzer

Figure 6.5: Usecase Diagram

6.2.3 Sequence Diagram

The sequence diagram is used primarily to show the interactions between objects
in the sequential order that those interactions occur. Developers typically think sequence
diagrams were meant exclusively for them. However, an organization’s business staff can
find sequence diagrams useful to communicate how the business currently works by showing
how various business objects interact.Sequence diagrams illustrate how objects interact with
each other. They focus on message sequences, that is, how messages are sent and received

30
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Doctor Prescription Analyzer

between a number of objects. The main purpose of sequence diagram is to show the order
of events between the parts of system that are involved in particular interaction.

Figure 6.6: Sequence Diagram

6.2.4 Class Diagram

The class diagram shows the building blocks of any object oriented system. Class
diagram depicts a static view of the model or part of the model, describing what attributes
and behavior it has rather that the detailing the methods of achieving operations. Class
diagrams are most useful in illustrating relationships between classes and interfaces. Gener-
alizations, aggregations, and associations are all valuable in reflecting interface, composition
or usage and connections receptively.
The Figure 6.2 illustrates aggregation relationships between classes. The lighter

31
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Doctor Prescription Analyzer

aggregation indicates that the class ObjectExplorer used ThumbNail, but does not neces-
sarily contain an instance of it.The strong, composite aggregations by the other connectors
indicate ownership or containment of the source classes by the target. Class, for example
Video Player values will be contained in Table Of Contents.

6.3 Summary

Thus we saw the various modeling techniques used for the design of M-Learning
Framework and the M-Learning Applications.

32
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Chapter 7

METHODOLOGY AND SYSTEM DESIGN

7.1 SYSTEM ARCHITECTURE

The architecture of the Machine Learning-based Doctor Receipt Analyzer system is de-
signed to seamlessly integrate OCR technology and machine learning algorithms, providing
an intelligent solution for efficient processing of medical receipts. At its core, the system
follows a modular architecture, comprising distinct components that collectively contribute
to its functionality. The user interface serves as the entry point, allowing healthcare pro-
fessionals to upload receipt images for analysis. The OCR engine, powered by a library
such as Tesseract, converts the receipt images into machine-readable text. The parsed text,
containing essential information, is then processed by machine learning algorithms to ex-
tract key details such as patient names, dates, services provided, and associated costs. The
extracted data is securely stored in a relational database for future retrieval and analy-
sis. A reporting and analytics module enables users to generate insights and reports from
the processed receipt data. The system also incorporates security measures, including en-
cryption and access controls, to ensure compliance with healthcare data privacy standards.
This modular and integrated architecture empowers the Machine Learning-based Doctor
Receipt Analyzer system to streamline administrative workflows, reduce manual efforts,
and enhance the accuracy of medical receipt processing in healthcare environments.

33
 Doctor Prescription Analyzer

Figure 7.1: System Architecture

34
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Doctor Prescription Analyzer

7.2 SYSTEM MODULES

7.2.1 User Interface Module:

This module provides a user-friendly web interface for healthcare professionals to interact
with the system. It includes features for user authentication, receipt uploads, and access to
system functionalities.

7.2.2 Image Preprocessing Module:

Responsible for optimizing receipt images before OCR processing. This module ensures
that images are standardized, enhancing OCR accuracy.

7.2.3 OCR Processing Module:

Integrates Optical Character Recognition (OCR) technology (e.g., Tesseract) to convert

preprocessed receipt images into machine-readable text.
AI Engine: The AI engine manages model training, updates, and optimization. Text
Extraction and Parsing Module: Processes the OCR-converted text, extracting relevant
information such as patient names, dates, services provided, and costs. This module employs
machine learning algorithms for pattern recognition and data extraction.

35
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Chapter 8

PROJECT IMPLEMENTATION

8.1 OVERVIEW OF PROJECT MODULES

1. An intuitive web-based interface facilitates seamless interaction for healthcare pro-

fessionals, allowing easy upload and processing of receipts. Enhances user experience
and accessibility, reducing the learning curve for system utilization.

2. Integrates advanced OCR technology (e.g., Tesseract) to accurately extract text from
scanned or photographed receipt images. Enables automated conversion of images
into machine-readable text, reducing manual data entry efforts.

3. Incorporates image preprocessing techniques, including resizing, normalization, and

contrast adjustments, to optimize OCR accuracy. Enhances the quality of receipt
images for improved OCR performance.

4. Utilizes machine learning algorithms for intelligent data extraction, parsing key in-
formation such as patient names, dates, services provided, and associated costs.

5. Stores extracted data in a secure relational database (e.g., MySQL or PostgreSQL)

for organized and efficient data retrieval. Facilitates seamless access to processed data
and ensures secure and structured storage.

6. Provides reporting and analytics features, allowing users to generate insights, trends,
and reports from the processed receipt data. Enables data-driven decision-making,
contributing to improved financial and operational management.

36
 Doctor Prescription Analyzer

8.2 TOOLS AND TECHNOLOGY USED

8.2.1 OCR

OCR, or Optical Character Recognition, is a technology that converts different types of

documents, such as scanned paper documents, PDFs, or images captured by a digital cam-
era, into editable and searchable data. It recognizes text characters and converts them into
machine-readable text.
Key Components and Processes of OCR Technology:

• Image Acquisition: The process begins with the acquisition of an image containing
text. This can be a scanned document, a photograph, or a digital image.

• Preprocessing: Description: The acquired image may undergo preprocessing to en-

hance the quality of the text. This involves tasks like noise reduction, contrast ad-
justment, and image normalization.

• Text Localization: OCR technology identifies and locates the areas of the image that
contain text. This step is crucial for isolating the text for recognition.

• Character Recognition: In this core step, OCR software recognizes individual charac-
ters within the localized text regions. It matches the shapes of the characters against
a database of known fonts and shapes.

• Post-processing: After character recognition, post-processing may be applied to refine

and improve the accuracy of the recognized text. This may involve context analysis
and correction of errors.

• Output: The final output of OCR is machine-readable text that can be edited,
searched, or stored digitally. This output can be used for various applications, such
as document indexing, data extraction, and text analysis.

37
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Doctor Prescription Analyzer

8.3 System Algorithm

1. Data Input and Integration:

- Healthcare providers input patient data, including symptoms, medical history, and
current medications, into the system through the user interface.
- The system integrates with electronic health records (EHRs) to retrieve comprehen-
sive patient information, including previous diagnoses and prescriptions.

2. Natural Language Processing (NLP):

- NLP algorithms process and analyze free-text inputs, extracting relevant information
from patient descriptions of symptoms and medical history.

3. Data Preprocessing:
- Preprocessing techniques are applied to clean and structure the data, including text
normalization, tokenization, and removal of stopwords and irrelevant information.

4. Machine Learning for Symptom Analysis:

- Machine learning models analyze the preprocessed symptom data, taking into ac-
count the patient’s medical history, to generate a list of potential diagnoses and
associated probabilities.

5. Drug Database Query:

- The system queries a comprehensive drug database to retrieve information on avail-
able medications, dosages, side effects, and potential interactions.

6. Personalized Drug Recommendations:

- The machine learning algorithms consider factors such as the patient’s medical
history, allergies, and potential drug interactions to generate personalized medication
recommendations based on the diagnosed condition(s).

7. Clinical Guidelines Integration:

- The system aligns medication recommendations with established clinical guidelines
and best practices to ensure that prescribed medications meet medical standards.

8. Prescription Analysis:
- Concurrently, the system evaluates doctor prescriptions in real-time, comparing
them to clinical guidelines, identifying potential drug interactions, and optimizing
the medication choices for individual patients.

38
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Doctor Prescription Analyzer

9. Security and Privacy Checks:

- Robust security measures are applied to protect patient data and ensure compliance
with healthcare privacy regulations, maintaining the confidentiality and integrity of
sensitive information throughout the process.

10. Feedback Mechanism:

- Healthcare providers have the option to provide feedback on medication recommen-
dations and prescription analyses, contributing to the system’s continuous improve-
ment.

39
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Chapter 9

DEPLOYMENT AND MAINTENANCE

9.1 Deployment

• Infrastructure Planning: Plan the infrastructure required for deploying the system,
including server specifications, database management, and networking considerations.

• Software Installation: Install the necessary software components, including the web
application, database management system (DBMS), OCR libraries, and any other
dependencies.

• Configuration Settings: Configure the system settings, including OCR library param-
eters, database connections, and user authentication settings. Ensure compatibility
with the intended deployment environment.

• Security Setup: Implement security measures, such as encryption protocols, access

controls, and secure transmission, to protect sensitive patient and financial data.
Ensure compliance with healthcare data privacy standards.

• Testing: Conduct thorough testing of the deployed system to identify and address
any issues related to functionality, performance, and security.

9.2 Deployment

• Regular Updates: Implement regular updates to the system, including security patches,
bug fixes, and feature enhancements. Utilize version control systems for efficient up-
date management.

• Monitoring and Performance Optimization: Monitor system performance, identify-

ing bottlenecks or issues. Optimize the system for efficiency, ensuring smooth and
responsive user interactions.

40
 Doctor Prescription Analyzer

• Data Backup and Recovery: Implement regular data backup procedures to safeguard
against data loss. Establish a robust data recovery plan in case of unexpected events.

• Security Audits: Conduct periodic security audits to assess the system’s vulnerability
to potential threats. Address any identified security gaps to maintain data integrity
and confidentiality.

• User Support and Feedback: Provide ongoing user support to address queries or
issues. Collect user feedback to identify areas for improvement and enhance user
satisfaction.

• Compliance Checks: Regularly review and update the system to ensure continued
compliance with healthcare industry regulations, such as HIPAA, and any other rel-
evant standards.

41
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Chapter 10

PROJECT ESTIMATION,SCHEDULE AND TEAM

STRUCTURE

10.1 Project Estimation

Here the prediction is made about the size of total project. Effective software project
estimation is one of the most challenging and important activity in software development
once you have on estimate size of your product you can desire the effort estimate.

10.2 Reconciled Estimates

Cost Estimate

Like all estimation model, the COCOMO model requires sizing information. This informa-
tion can be specified in the form of
1. Object Point (OP)
2. Function Point (FP)
3. Lines of Source Code (KLOC)
For our project, we use the sizing information in the form of Lines of source code.

42
 Doctor Prescription Analyzer

10.2.1 Efforts

Equation for calculation of effort in person-month for the COCOMO model is:
E = a * (KLOC) bWhere;
a=3.2
b=1.05, for semi-detached projects
E=Effort in person-months
D = E/N
Where,
E=Effort in person-months
N=Number of persons required
D=Duration of project in months.

10.2.2 Development time per month

E=3.2(KLOC)1:05E = 3:2(4:2)1:05
E=4*30 Person-month
Development time:
D=E/N
D=4*30/4
D=3.82 month

10.2.3 Development time for Project

Requirements analysis require 3 months

Implementation and testing requires 3.82 months.
Total Duration for completion of project D= 6.82 months.

10.2.4 Number of Persons

Total Four persons are required to complete the project successfully within given time span.

43
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Doctor Prescription Analyzer

10.2.5 Project Resources

Hardware Resources Required Software Resources Required

Table 10.1: Hardware Resources Required

Sr,No Parameter Minimum Requirement

1 Processor Pentium IV/Intel i3 core
2 Speed 1.1GHZ
3 Ram 512MB
4 Hard Disk 20GB
5 Keyboard Standard keyword
6 Mouse Two or Three Button
7 Monitor LED Monitor

Platform :
1. Operating System : Windows XP / 7
2. Programming Language : Python
3. Software Version : 3.10 or above
4. Tools : Notepad ++ , Vscode
5. Front End : Boot Strap
6. Data Base : mysql or firebase

44
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Doctor Prescription Analyzer

10.3 Risk Management w.r.t. NP Hard analysis

• Np Hard Problem : In our project we have face issued like syntax errors , code error
which is easy solvable.

8888.png

Figure 10.1: NP-hard and NP-complete

45
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Doctor Prescription Analyzer

10.3.1 Risk Identification

For risks identification, review of scope document, requirements specifications and schedule
is done. Answers to questionnaire revealed some risks. Each risk is categorized as per the
categories mentioned in. Please refer table 2.1 for all the risks. You can refer following risk
identification questionnaire.

1 Have top software and customer managers formally committed to support the project?

2 Are end-users enthusiastically committed to the project and the system/product to

be built?

3 Are requirements fully understood by the software engineering team and its cus-
tomers?

4 Have customers been involved fully in the definition of requirements?

5 Do end-users have realistic expectations?

6 Does the software engineering team have the right mix of skills?

7 Are project requirements stable?

8 Is the number of people on the project team adequate to do the job?

9 Do all customer/user constituencies agree on the importance of the project and on

the requirements for the system/product to be built?

10.3.2 Product Size Risk

R1 = Risk associated with the overall size of the software to be built or modified.
R2 = Project may not complete on time.
R3 = Product size may increase due to inefficient implementation of system.

10.3.3 Business Impact Rule

R4 = Risk associated with constraints imposed by management or the marketplace.

R5 = Product loses its market value.

46
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Doctor Prescription Analyzer

10.3.4 Process Risk

R7 = Risk associated with the degree to which the software process has been defined and
is followed by the development organization.
R8 = Software process model is not followed up to the define degree.
R9 = Selection of software process model.

10.3.5 Technical Risk

R10 = Risk associated with the complexity of the system to be built and the newness of
the technology that is packaged by the system.
R11 = System is not scalable.
R12 = System may fail to provide desired efficiency to customer

10.3.6 Development Environment Risk

R13 = Risk associated with the availability and quality of the tools to be used to build the
product.
R14 = Lack of training on development tools and inexperience.
R15 = It is difficult to alter the system as per customer’s requirements in later stage.

10.3.7 Staff Size And Experience Risk

R16 = Risk associated with the overall technical and project experience of the software
engineers who will do the work.

R17 = Less technical and project experience of the software engineers who are working
on project.

47
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Doctor Prescription Analyzer

10.3.8 Risk Analysis

The risks for the Project can be analyzed within the constraints of time and quality

Table 10.2: Risk Analysis

ID Risk Description Probability Schedule Quality Overall

1 Will the system process 100-150 system Low Low High High
Will the system able to handle
2 Low Low High High
multiple users at a time
3 Will the system handle multiple error? Low Low High High
Will system degrade the
4 Low Low High High
performance of other subsystem

Table 10.3: Probability and Value

Probability Value Critical

High Probability of occurrence is ¿75
Medium Probability of occurrence is 26 -75
Low Probability of occurrence is ¡25

Table 10.4: Impact and Value

Impact Value Description

Schedule
Very high >10
impact or unacceptable Quality
Schedule
High 5-10
impact or some parts of the project have low quality
Schedule
Medium <5 impact or barely noticeable degradation
in quality ow impact on schedule or quality can be incorporated

48
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Doctor Prescription Analyzer

10.4 Feasibility

The feasibility of the project is analysed in this phase and business proposal is put
forth with a very general plan for the project and some cost estimates. During system
analysis the feasibility study of the proposed system is to be carried out. This is to ensure
that the proposed system is not a burden to the company. For feasibility analysis, some
understanding of the major requirements for the system is essential.Dimensions of Software
Feasibility are as follows:

• Technology:
Is project technically feasible?
Is it within state of art?
Can defect be reduce to a level matching application’s need?

• Finance:
Is it financially feasible?
Can development be completed at a cost the software organization and its client or
market can afford?

• Time:
Will project’s time to market beat competition?

• Resources:
Does the organization have resources needed to success?

Two key considerations involved in the feasibility analysis are:

1. Technical Feasibility. 2. Cost Feasibility.

49
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Doctor Prescription Analyzer

10.4.1 Technical Feasibility

This study is carried out to check the technical feasibility, that is, the technical re-
quirements of the system. Any system developed must not have a high demand on the
available technical resources. This will lead to high demands on the available technical
resources. This will lead to high demands being placed on the client. The developed system
must have a modest requirement,as only minimal or null changes are required for imple-
menting this system. Technical feasibility assessment can be done through following ways:
1)NP-Complete. 2) NP-Hard. 3)Satisfiability.

10.4.2 Cost Feasibility

This study is carried out to check the economic impact that the system will have
on the organization. The amount of fund that the company can pour into the research
and development of the system is limited. The expenditures must be justified. Thus the
developed system as well within the budget and this was achieved because most of the
technologies used are freely available. Only the customized products had to be purchased.

10.5 Project Schedule

10.5.1 Project task set

Major Tasks in the Project stages are:

• Task 1: Requirement Analysis (Base Paper Explanation).

• Task 2: Project Specification (Paper Work).

• Task 3: Technology Study and Design.

• Task 4: Coding and Implementation (Module Development).

10.6 Project Plan

50
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Doctor Prescription Analyzer

Table 10.5: Planner and Progress Report for project

Phase Activity Start Date End Date Group Mem-

bers
1 Selection of Project Topic 22-08-2023 25-08-2023 Team
2 Functional Requirement 29-08-2023 09-09-2023 Team
Specification(FRS)
3 Design Prototype 11-09-2023 21-09-2023 Team
4 Set Theory and Math 23-09-2023 06-09-2023 team
Model
5 UML Diagram Prototype 23-09-2023 03-10-2023 Team
6 Project Problem Statement 08-10–2023 19-10-2023 team
using NP Complete
7 UML Diagram in StarUML 05-10-2023 22-10-2023 Team
8 Paper Presentation 05-11-2023 05-11-2023 Team
9 Software Requirement 6-11-2023 10-11-2023 Team
Specification
10 Paper Publication sem 1 11-11-2023 15-11-2023 Team
11 Report Submission Sem 1 15-11-2023 1-12-2023 Team
12 Module Design (admin, 1-12-2023 1-1-2023 Team
route, user)
13 GUI Design 1-11-2024 15-1-2024 Team
14 Arduino Programming 15-1-2024 1-2-2024 Team
15 Hardware Interfacing 1-2-2024 15-2-2024 Team
16 Project Implementation 15-2-2024 1-3-2024 Team
17 QA and Testing 1-3-2024 15-3-2024 Team
18 Outcome of Project 15-3-2024 15-4-2024 Team
19 Paper Publication sem 2 15-4-2024 25-4-2024 Team
19 Report Submission sem 2 25-4-2024 8-5-2024 Team

51
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Chapter 11

SOFTWARE TESTING AND VALIDATION

11.1 Introduction

Software testing is an activity aimed at evaluating an attribute or capability of a

program or system and determining that it meets its required results. It is more than just
running a program with the intention of finding faults. Every project is new with different
parameters. No single yardstick maybe applicable in all circumstances. This is a unique
and critical area with altogether different problems. Although critical to software quality
and widely deployed by programs and testers. Software testing steel remains an art, due
to limited understanding of principles of software. The difficulty stems from complexity
of software. The purpose of software testing can be quality assurance, verification and
validation or reliability estimation. Software testing is a trade-off between budget,time and
quality. In this chapter there is relevant explanation on testing strategies use to test the
system, and test cases.

11.2 Types of Testing

Testing Strategy used for testing the system are as follows,

1. Manual Testing
2. Automated Testing
3. Unit Testing
4. Integration Testing
5. Regression Testing

52
 Doctor Prescription Analyzer

11.2.1 Manual Testing

Manual and Automated test are the types of software testing. We are doing a manual
test for testing our system that is without using any automated tool or any script. In
this type tester takes over the role of an end user and test the software to identify any
unexpected behavior or bug. There are different stages for manual testing like unit testing,
integration testing, system testing and user acceptance testing. Testers use test plan, test
cases or test scenario to test the software to ensure the completeness of a testing. Manual
testing also includes exploratory testing as a testers explore the software to identify the
errors in it.

11.2.2 Automated Testing

Automation testing which is also known as Test Automation is when the tester writes
scripts and uses software to test the software. This process involves automation of a man-
ual process. Automation testing is used to re-run the test scenarios that were performed
manually, quickly and repeatedly.

11.2.3 Unit Testing

In case of unit testing, each software component, software modules or software

subsystem is tested independent of any other components involved in the whole software
system. That is individual software modules or software components are tested in unit
testing. The main agenda behind unit testing is to verify and validate each and every unit
of the software system by checking its working and performance and comparing it with the
software specification. The significant control paths are tested and verified to discover errors
within the boundary of the module and the component level design used for the same.

11.2.4 Integration Testing

Integration testing is a kind of testing meant for building the software architecture
along with finding out the errors related with the interfacing. After successful execution
of unit testing, software subsystem will be collected together and combined together in
order to build the whole software system as it is specified and define at high level design.
Facial Emotion Recognition Using Convolutional Neural Network Integration testing is an
efficient procedure for verification of the structure of a software system and validation of
order of execution of software system while conducting tests to determine errors allied with
interfacing.

53
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Doctor Prescription Analyzer

11.2.5 Regression Testing

During the software development procedure, whenever the software system is modified
by means of editing, removing, adding source code, software developers need to be sure that
the new version of the software is good as earlier version. Tests that focus on the software
modules that have been modified or altered and focus on overall functionality of the software
system when the software functions are likely to be affected by the modifications or change.

11.3 Software testing

Software testing can also provide an objective, independent view of the software to
allow the business to appreciate and understand the risks of software implementation. Test
techniques include the process of executing a program or application with the intent of
finding software bugs (errors or other defects), and verifying that the software product
is fit for use. Software testing involves the execution of a software component or system
component to evaluate one or more properties of interest. In general, these properties
indicate the extent to which the component or system under test:

• Meets the requirements that guided its design and development,

• Responds correctly to all kinds of inputs,

• Performs its functions within an acceptable time,

• It is sufficiently usable,

• Can be installed and run in its intended environments, and

• Achieves the general result its stakeholders desire.

As the number of possible tests for even simple software components is practically
infinite, all software testing uses some strategy to select tests that are feasible for the
available time and resources. As a result, software testing typically (but not exclusively)
attempts to execute a program or application with the intent of finding software bugs (errors
or other defects). The job of testing is an iterative process as when one bug is fixed, it can
illuminate other, deeper bugs, or can even create new ones.

54
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Doctor Prescription Analyzer

11.4 Black Box Testing

This testing methodology looks at what are the available inputs for an application and
what the expected outputs are that should result from each input. It is not concerned with
the inner workings of the application, the process that the application undertakes to achieve
a particular output or any other internal aspect of the application that may be involved in
the transformation of an input into an output. Most black-box testing tools employ either
coordinate based interaction with the applications graphical user interface (GUI) or image
recognition. An example of a black-box system would be a search engine. You enter text
that you want to search for in the search bar, press “Search” and results are returned to
you. In such a case, you do not know or see the specific process that is being employed to
obtain your search results, you simply see that you provide an input – a search term – and
you receive an output – your search results.

11.4.1 Black-box

There are many advantages to black-box testing. Here are a few of the most commonly cited:

1. Ease of use: Because testers do not have to concern themselves with the inner workings
of an application, it is easier to create test cases by simply working through the application,
as would an end user.

2. Quicker test case development: Because testers only concern themselves with
the GUI, they do not need to spend time identifying all of the internal paths that may be
involved in a specific process, they need only concern themselves with the various paths
through the GUI that a user may take.

3. Simplicity: Where large, highly complex applications or systems exist black-box test-
ing offers a means of simplifying the testing process by focusing on valid and invalid inputs
and ensuring the correct outputs are received.
But, for all of the benefits of black-box testing, many attempts to create black-box test
systems resulted in several drawbacks that caused people to question the viability of the
black-box approach.

Some of the most commonly cited issues were:

55
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Doctor Prescription Analyzer

1. Script maintenance: While an image-based approach to testing is useful, if the

user interface is constantly changing the input may also be changing. This makes script
maintenance very difficult because black-box tools are reliant on the method of input being
known.

2. Fragility: Interacting with the GUI can also make test scripts fragile. This is be-
cause the GUI may not be rendered consistently from time to time on different platforms
or machines. Unless the tool is capable of dealing with differences in GUI rendering, it is
likely that test scripts will fail to execute properly on a consistent basis.

3. Lack of introspection: Ironically, one of the greatest criticism of black-box testing is

that it isn’t more like white-box testing; it doesn’t know how to look inside an application
and therefore can never fully test an application or system. The reasons cited for needing
this capability are often to overcome the first two issues mentioned. The reality is quite
different.

56
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Doctor Prescription Analyzer

11.5 White Box Testing

This testing methodology looks under the covers and into the subsystem of an applica-
tion. Whereas black-box testing concerns itself exclusively with the inputs and outputs of an
application, white-box testing enables you to see what is happening inside the application.
White box testing provides a degree of sophistication that is not available with black-box
testing as the tester is able to refer to and interact with the objects that comprise an ap-
plication rather than only having access to the user interface. An example of a white-box
system would be in-circuit testing where someone is looking at the interconnections between
each component and verifying that each internal connection is working properly. Another
example from a different field might be an auto-mechanic who looks at the inner-workings
of a car to ensure that all of the individual parts are working correctly to ensure the car
drives properly.

11.5.1 White-box

Like black-box testing, there are distinct advantages to white-box testing. Here are a few
of the most commonly cited:

1. Introspection: Introspection, or the ability to look inside the application, means that
testers can identify objects programmatically. This is helpful when the GUI is changing
frequently or the GUI is yet unknown as it allows testing to proceed. It also can, in some sit-
uations, decrease the fragility of test scripts provided the name of an object does not change.

2. Stability: In reality, a by-product of introspection, white-box testing can deliver

greater stability and reusability of test cases if the objects that comprise an application
never change.

3. Thoroughness: In situations where it is essential to know that every path has been
thoroughly tested, that every possible internal interaction has been examined, white-box
testing is the only viable method.

As such, white-box testing offers testers the ability to be more thorough in terms of how
much of an application they can test. Despite these benefits, white-box testing has its
drawbacks.

57
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Doctor Prescription Analyzer

Some of the most commonly cited issues are:

1. Complexity: Being able to see every constituent part of an application means that a
tester must have detailed programmatic knowledge of the application in order to work with
it properly. This high-degree of complexity requires a much more highly skilled individual
to develop test case.

2. Fragility: While introspection is supposed to overcome the issue of application changes

breaking test scripts the reality is that often the names of objects change during product de-
velopment or new paths through the application are added. The fact that white-box testing
requires test scripts to be tightly tied to the underlying code of an application means that
changes to the code will often cause white-box test scripts to break. This, then, introduces
a high degree of script maintenance into the testing process.

3. Integration: For white-box testing to achieve the degree of introspection required

it must be tightly integrated with the application being tested. This creates a few problems.
To be tightly integrated with the code you must install the white-box tool on the system on
which the application is running. This is okay, but where one wishes to eliminate the possi-
bility that the testing tool is what is causing either a performance or operational problem,
this becomes impossible to resolve. Another issue that arises is that of platform support.
Due to the highly integrated nature of white-box testing tools many do not provide support
for more than one platform, usually Windows®. Where companies have applications that
run on other platforms, they either need to use a different tool or resort to manual testing.

58
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Chapter 12

RESULT AND ANALYSIS

12.1 System Layout

59
 Doctor Prescription Analyzer

60
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Doctor Prescription Analyzer

61
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Doctor Prescription Analyzer

62
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Doctor Prescription Analyzer

63
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Doctor Prescription Analyzer

64
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Chapter 13

ADVANTAGES LIMITATIONS AND APPLICATIONS

Advantages

1. Automation of Manual Processes: The system automates the traditionally manual

and time-consuming task of processing doctor receipts, reducing the need for manual
data entry and streamlining administrative workflows.

2. Improved Accuracy in Data Extraction: Machine learning algorithms enhance the

accuracy of data extraction from receipts, even when dealing with handwritten or
complex formats.

3. Versatility in Handling Diverse Receipt Formats: The system is designed to handle

various receipt formats, including both printed and handwritten text, ensuring adapt-
ability to the diverse documentation styles used across different healthcare providers.

4. Enhanced Financial Management: The system contributes to more effective financial

management within healthcare institutions by providing a reliable and automated
tool for processing and organizing receipt data.

5. Integration into Healthcare Workflows: Seamlessly integrates into existing healthcare

workflows, ensuring minimal disruption to established processes.

6. Time and Resource Savings: The automation of receipt processing saves time and
resources by reducing the manual effort required for data entry and record-keeping.

7. Improvement in Patient Care: By automating administrative tasks, the system allows

healthcare professionals to focus more on patient care responsibilities, improving the
overall quality of healthcare services.

65
 Doctor Prescription Analyzer

Applications

1. Medical Billing Automation: Automates the extraction of billing information from

doctor receipts, accelerating the billing process and reducing the need for manual
data entry.

2. Healthcare Analytics: Utilizes the processed receipt data for healthcare analytics,
providing insights into spending patterns, service utilization, and financial perfor-
mance.

3. Patient Record Integration: Integrates receipt data into patient records, creating a
comprehensive view of financial transactions associated with patient care.

4. Research and Audit Support: Facilitates research activities and audit processes by
providing accurate and organized receipt data for analysis.

5. Research and Audit Support: Facilitates research activities and audit processes by
providing accurate and organized receipt data for analysis.

6. Enhanced Financial Transparency: Provides a transparent view of financial transac-

tions, contributing to improved financial transparency within healthcare institutions.

66
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Doctor Prescription Analyzer

Future Scope

Future iterations may focus on seamless integration with EHR systems, providing a unified
platform for healthcare professionals to access both medical and financial information in one
interface. Continued advancements in machine learning algorithms may lead to improved
accuracy in data extraction, especially in handling diverse receipt formats, handwritten
text, and complex document structures.

67
Department of Computer Engineering, SVIT,Nashik, 2023-24
 Chapter 14

SUMMARY AND CONCLUSION

In conclusion, the Machine Learning-based Doctor Receipt Analyzer system harnessing

OCR technology marks a significant leap forward in healthcare administration. This innova-
tive solution has successfully automated and streamlined the intricate process of processing
doctor receipts, transcending the limitations posed by diverse formats and handwritten en-
tries. By seamlessly integrating machine learning algorithms, the system not only enhances
the accuracy and efficiency of data extraction but also promises to revolutionize financial
management and decision-making within healthcare institutions. Its adaptability, commit-
ment to data security, and potential for future advancements position it as a pivotal tool
in the evolving landscape of healthcare administration. As we embrace this cutting-edge
technology, we anticipate a future where administrative workflows are further optimized,
allowing healthcare professionals to allocate more time and resources to the core of their
practice – delivering exceptional patient care.

68
 Bibliography

[1] Telemedicine, https://www.mohfw.gov.in/pdf/Telemedicine.pdf

[2] Wittich CM, Burkle CM, Lanier WL. Medication errors: an overview for clinicians.
Mayo Clin Proc. 2014 Aug;89(8):1116-25.

[3] CHEN, M. R., WANG, H. F. (2013). The reason and prevention of hospital medication
errors. Practical Journal of Clinical Medicine,

[4] Drug Review Dataset, https://archive.ics.uci.edu/ml/datasets/Drug

[5] Fox, Susannah, and Maeve Duggan. ”Health online 2013. 2013.”

[6] Bartlett JG, Dowell SF, Mandell LA, File TM Jr, Musher DM, Fine MJ. Practice guide-
lines for the management of community-acquired pneumonia in adults. Infectious Dis-
eases Society of America. Clin Infect Dis. 2000 Aug;31(2):347-82. doi: 10.1086/313954.
Epub 2000 Sep 7. PMID: 10987697; PMCID: PMC7109923.

[7] Fox, Susannah Duggan, Maeve. (2012). Health Online 2013. Pew Research Internet
Project Report.

[8] T. N. Tekade and M. Emmanuel, ”Probabilistic aspect mining approach for interpre-
tation and evaluation of drug reviews,” 2016 International Conference on Signal Pro-
cessing, Communication, Power and Embedded System (SCOPES), Paralakhemundi,
2016, pp. 1471-1476, doi: 10.1109/SCOPES.2016.7955684.

[9] Doulaverakis, C., Nikolaidis, G., Kleontas, A. et al. GalenOWL: Ontology-based drug
recommendations discovery. J Biomed Semant 3, 14 (2012).

[10] Leilei Sun, Chuanren Liu, Chonghui Guo, Hui Xiong, and Yanming Xie. 2016. Data-
driven Automatic Treatment Regimen Development and Recommendation.

69
 Appendix A

DETAILS OF THE PAPERS PUBLICATION

70
 Analyze Doctor Receipt with Optical Character
Recognition Technology And Machine Learning
1
Patait Sumit, 2Patil Sudarshan, 3Patil Kalpesh, 4Avhad Roshan,
5
Prabha Gadakh
Department Of Computer Engineering
Sir Visvesvaraya Institute Of Technology
Nashik, Maharashtra, India

Abstract- In this innovative application, we present a Machine Learning-based Doctor Receipt

Analyzer system empowered by Optical Character Recognition (OCR) technology. The system
represents a transformative leap in the processing of medical receipts by seamlessly integrating
advanced OCR algorithms. Capable of extracting critical information from scanned or photographed
receipts, the solution automates the traditionally manual task of deciphering handwritten or printed
text, enhancing both accuracy and efficiency. Tailored for effortless integration into healthcare
administration, the system intelligently decodes essential details such as patient names, dates, services
provided, and associated costs. The synergy of machine learning and OCR not only streamlines billing
processes for healthcare professionals but also promises to elevate the overall efficiency of medical
documentation and financial management, marking a significant advancement in healthcare
technology.

Keywords: OCR, receipt analyzer, Machine Learning, NLP.

Published in IJIRMPS (E-ISSN: 2349-7300), Volume 11, Issue 6, Nove-Dec 2023

License: Creative Commons Attribution-ShareAlike 4.0 International License

INTRODUCTION
Introducing a cutting-edge solution, our Machine Learning-based Doctor Receipt Analyzer system harnesses
the power of OCR technology to revolutionize the way medical receipts are processed. By leveraging
sophisticated Optical Character Recognition algorithms, the system adeptly extracts crucial information from
scanned or photographed receipts. This innovative approach not only automates the traditionally manual and
time-consuming task of deciphering handwritten or printed text but also enhances accuracy and efficiency.
Seamlessly integrating into healthcare administration, the system intelligently decodes patient names, dates,
services provided, and associated costs. The synergy of machine learning and OCR not only streamlines the
billing and record-keeping process for healthcare professionals but also promises to elevate the overall
efficiency of medical documentation and financial management.
This groundbreaking Machine Learning-driven Doctor Receipt Analyzer signifies a paradigm shift in
healthcare management. By seamlessly merging the capabilities of OCR technology with advanced
algorithms, our system revolutionizes the processing of medical receipts. Its prowess lies in its ability to
adeptly extract vital information from diverse receipt formats, be they scanned or photographed. Beyond
automating the traditionally labor-intensive task of deciphering handwritten or printed text, the system
significantly boosts accuracy and operational efficiency. Tailored to seamlessly integrate into healthcare
administration workflows, it intelligently decodes key elements such as patient names, dates, services
rendered, and associated costs. This harmonious blend of machine learning and OCR not only simplifies
billing and record-keeping for healthcare professionals but also holds the promise of elevating the overall
efficiency of medical documentation and financial management to unprecedented levels.

IJIRMPS230385 Website: www.ijirmps.org Email: editor@ijirmps.org 1

LITURATURE SURVEY
• This paper utilizes the existing recommendation models either give only one recommendation
(however, there may be a variety of drug combination options in practice) or cannot provide the confidence
level of the recommended result. To fill these gaps, a Drug Recommendation model based on Message
Propagation neural network (denoted as DRMP) is proposed in this paper. Then, the Drug-Drug Interaction
(DDI) knowledge is introduced into the proposed model to reduce the DDI rate in recommended drugs.
Finally, the proposed model is extended to Bayesian Neural Network (BNN) to realize multiple
recommendations and give the confidence of each recommendation result, so as to provide richer information
to help doctors make decisions. Experimental results on public data sets show that the proposed model is
superior to the best existing models. [1].
• This study builds a drug recommendation system for patients who have diabetes. One method applied
in a recommendation is Case Base Reasoning (CBR). The CBR method calculates based on the similarity of
the old case to the new issue. In calculating the similarity of the problems using the Nearest Neighbors
algorithm. Although there have been many studies on drug recommendations, no studies have added risk
factor variables. Risk factors can influence results more accurately. The results of this study are in the form
of a prescription drug recommendation system for diabetic patients based on the type of diabetes they suffer
from. They are testing the accuracy of this system using the Confusion Matrix with an accuracy of 80.60\%.
Suggestions for further research are to add several types of similarity. It is necessary to research different
kinds of diseases. It is expected to produce an integrated system that can provide drug recommendations for
various conditions. [2].
• In this paper, we design and implement a drug recommender system framework that applies sentiment
analysis technologies on drug reviews. The objective of this research is to build a decision-making support
platform to help patients to achieve more significant choices in drug selection. Firstly, we propose a
sentimental measurement approach to drug reviews and generate ratings on drugs. Secondly, we take how
much the drug reviews are useful to users, patient's conditions, and dictionary sentiment polarity of drug
reviews into consideration. Then, we fuse those factors into the recommendation system to list appropriate
medications. Experiments have been carried out using Decision Tree, K-Nearest Neighbors, and Linear
Support Vector Classifier algorithm in rating generation and Hybrid model in recommendation based on the
given open dataset. The analysis is carried out to tune the parameters for each algorithm in order to achieve
greater performance. Finally, Linear Support Vector Classifier is selected for rating generation to obtain a
good trade-off among model accuracy, model efficiency, and model scalability. [3].
• This research focuses on prescriptions of GuiZhi Decoction, which is from Zhongjing Zhang's
Treatise on Cold Pathogenic Diseases. First, we constructed the database for prescriptions of GuiZhi
Decoction based on the prescriptions' names, the concept extraction and formal expression. Then we generate
the formal context and optimized it by use of the mathematical theory of formal concept analysis and the
relationship between prescriptions and herbs in database. In the formal context, the prescriptions are objects
and the herbs are attributes. Next the structural partial-ordered attribute diagram is generated, so as to
visualize the knowledge of prescriptions of GuiZhi Decoction. Finally, through the analysis of structural
partial-ordered attribute diagram, more comprehensive, objective and multileveled knowledge are
discovered. [4].

AIM & OBJECTIVES

• To Eliminate manual efforts involved in extracting information from doctor receipts by employing
OCR to recognize and extract text seamlessly.
• To Design the system to be adaptable to various receipt formats, accommodating the diversity in
documentation styles used across different healthcare providers.

MOTIVATION
The motivation behind developing a Machine Learning-based Doctor Receipt Analyzer system using OCR
technology stems from the desire to address significant challenges in the current healthcare administration
landscape. Manual processing of doctor receipts is time-consuming and prone to errors. The motivation is to
enhance the efficiency of healthcare administration by automating the extraction of crucial information from

IJIRMPS230385 Website: www.ijirmps.org Email: editor@ijirmps.org 2

receipts, reducing the burden on healthcare professionals. Handwritten or varied receipt formats can pose
challenges in accurate data extraction.

SYSTEM ARCHITECTURE

Fig -1: System Architecture Diagram

DETAILS OF MODULE
• User Interface Module: This module provides a user-friendly web interface for healthcare
professionals to interact with the system. It includes features for user authentication, receipt uploads, and
access to system functionalities.
• Image Preprocessing Module: Responsible for optimizing receipt images before OCR processing.
This module ensures that images are standardized, enhancing OCR accuracy.
• Text Extraction and Parsing Module: Processes the OCR-converted text, extracting relevant
information such as patient names, dates, services provided, and costs. This module employs machine learning
algorithms for pattern recognition and data extraction.

APPLICATION
• Patient Record Integration: Integrates receipt data into patient records, creating a comprehensive view
of financial transactions associated with patient care.
• Research and Audit Support: Facilitates research activities and audit processes by providing accurate
and organized receipt data for analysis.
• Research and Audit Support: Facilitates research activities and audit processes by providing accurate
and organized receipt data for analysis.
• Enhanced Financial Transparency: Provides a transparent view of financial transactions, contributing
to improved financial transparency within healthcare institutions.

ALGORITHM
1. Data Input and Integration: The system integrates with electronic health records (EHRs) to retrieve
comprehensive patient information, including previous diagnoses and prescriptions.
2. Natural Language Processing (NLP): NLP algorithms process and analyze free-text inputs,
extracting relevant information from patient descriptions of symptoms and medical history.
3. Data Preprocessing: Preprocessing techniques are applied to clean and structure the data, including
text normalization, tokenization, and removal of stop words and irrelevant information.
4. Machine Learning for Symptom Analysis: Machine learning models analyze the preprocessed
symptom data, taking into account the patient's medical history, to generate a list of potential diagnoses and
associated probabilities.

IJIRMPS230385 Website: www.ijirmps.org Email: editor@ijirmps.org 3

5. Drug Database Query: The system queries a comprehensive drug database to retrieve information
on available medications, dosages, side effects, and potential interactions.
6. Personalized Drug Recommendations: The machine learning algorithms consider factors such as
the patient's medical history, allergies, and potential drug interactions to generate personalized medication
recommendations based on the diagnosed condition.
7. Clinical Guidelines Integration: The system aligns medication recommendations with established
clinical guidelines and best practices to ensure that prescribed medications meet medical standards.
8. Prescription Analysis: Concurrently, the system evaluates doctor prescriptions in real-time,
comparing them to clinical guidelines, identifying potential drug interactions, and optimizing the medication
choices for individual patients.

CONCLUSION
In conclusion, the Machine Learning-based Doctor Receipt Analyzer system harnessing OCR technology
marks a significant leap forward in healthcare administration. This innovative solution has successfully
automated and streamlined the intricate process of processing doctor receipts, transcending the limitations
posed by diverse formats and handwritten entries. By seamlessly integrating machine learning algorithms, the
system not only enhances the accuracy and efficiency of data extraction but also promises to revolutionize
financial management and decision-making within healthcare institutions. Its adaptability, commitment to
data security, and potential for future advancements position it as a pivotal tool in the evolving landscape of
healthcare administration. As we embrace this cutting-edge technology, we anticipate a future where
administrative workflows are further optimized, allowing healthcare professionals to allocate more time and
resources to the core of their practice – delivering exceptional patient care.

REFERENCES:
[1] Telemedicine, https:/ /www. mohfw. gov.in /pdf/ Telemedicine.pdf
[2] Wittich CM, Burkle CM, Lanier WL. Medication errors: an overview for clinicians. Mayo Clin Proc.
2014 Aug;89(8):1116-25.
[3] CHEN, M. R., & WANG, H. F. (2013). The reason and prevention of hospital medication errors. Practical
Journal of Clinical Medicine,
[4] Drug Review Dataset, https:// archive. ics. uci. edu/ ml/ datasets/Drug% 2B Review% 2B Dataset%
2B%2528 Drugs.com% 2529#.
[5] Fox, Susannah, and Maeve Duggan. ”Health online 2013. 2013.”
[6] Bartlett JG, Dowell SF, Mandell LA, File TM Jr, Musher DM, Fine MJ. Practice guidelines for the
management of community-acquired pneumonia in adults. Infectious Diseases Society of America. Clin
Infect Dis. 2000 Aug;31(2):347-82. doi: 10.1086/313954. Epub 2000 Sep 7. PMID: 10987697; PMCID:
PMC7109923.
[7] Fox, Susannah & Duggan, Maeve. (2012). Health Online 2013. Pew Research Internet Project Report.
[8] T. N. Tekade and M. Emmanuel, ”Probabilistic aspect mining approach for interpretation and evaluation
of drug reviews,” 2016 International Conference on Signal Processing, Communication, Power and
Embedded System (SCOPES), Paralakhemundi, 2016, pp. 1471-1476, doi:
10.1109/SCOPES.2016.7955684.
[9] Doulaverakis, C., Nikolaidis, G., Kleontas, A. et al. GalenOWL: Ontology-based drug recommendations
discovery. J Biomed Semant 3, 14 (2012).
[10] Leilei Sun, Chuanren Liu, Chonghui Guo, Hui Xiong, and Yanming Xie. 2016. Data-driven Automatic
Treatment Regimen Development and Recommendation.

IJIRMPS230385 Website: www.ijirmps.org Email: editor@ijirmps.org 4

Volume 12 Issue 2 @ 2024 IJIRMPS | ISSN: 2349-7300

Analyze Doctor Receipt with Optical Character

Recognition Technology and Machine Learning
1
Patait Sumit, 2Patil Sudarshan, 3Patil Kalpesh, 4Avhad Roshan,
5
Prabha Gadakh
Department Of Computer Engineering
Sir Visvesvaraya Institute Of Technology
Nashik, Maharashtra, India.

Abstract-
In this innovative application, we present a Machine Learning-based Doctor Receipt Analyzer system
empowered by Optical Character Recognition (OCR) technology. The system represents a
transformative leap in the processing of medical receipts by seamlessly integrating advanced OCR
algorithms. Capable of extracting critical information from scanned or photographed receipts, the
solution automates the traditionally manual task of deciphering handwritten or printed text,
enhancing both accuracy and efficiency. Tailored for effortless integration into healthcare
administration, the system intelligently decodes essential details such as patient names, dates, services
provided, and associated costs. The synergy of machine learning and OCR not only streamlines
billing processes for healthcare professionals but also promises to elevate the overall efficiency of
medical documentation and financial management, marking a significant advancement in healthcare
technology.

Keywords: OCR, receipt analyzer, Machine Learning, NLP.

Published in IJIRMPS (E-ISSN: 2349-7300), Volume 12, Issue 2, March- April 2024

License: Creative Commons Attribution-ShareAlike 4.0 International License

INTRODUCTION
Introducing a cutting-edge solution, our Machine Learning-based Doctor Receipt Analyzer system
harnesses the power of OCR technology to revolutionize the way medical receipts are processed. By
leveraging sophisticated Optical Character Recognition algorithms, the system adeptly extracts crucial
information from scanned or photographed receipts. This innovative approach not only automates the
traditionally manual and time-consuming task of deciphering handwritten or printed text but also enhances
accuracy and efficiency. Seamlessly integrating into healthcare administration, the system intelligently
decodes patient names, dates, services provided, and associated costs. The synergy of machine learning and
OCR not only streamlines the billing and record-keeping process for healthcare professionals but also
promises to elevate the overall efficiency of medical documentation and financial management.
This groundbreaking Machine Learning-driven Doctor Receipt Analyzer signifies a paradigm shift in
healthcare management. By seamlessly merging the capabilities of OCR technology with advanced
algorithms, our system revolutionizes the processing of medical receipts. Its prowess lies in its ability to
adeptly extract vital information from diverse receipt formats, be they scanned or photographed. Beyond
automating the traditionally labor-intensive task of deciphering handwritten or printed text, the system
significantly boosts accuracy and operational efficiency. Tailored to seamlessly integrate into healthcare
administration workflows, it intelligently decodes key elements such as patient names, dates, services
rendered, and associated costs. This harmonious blend of machine learning and OCR not only simplifies
billing and record-keeping for healthcare professionals but also holds the promise of elevating the overall
efficiency of medical documentation and financial management to unprecedented levels.

IJIRMPS230543 Website: www.ijirmps.org Email: editor@ijirmps.org 1

 Volume 12 Issue 2 @ 2024 IJIRMPS | ISSN: 2349-7300

LITURATURE SURVEY
• This paper utilizes the existing recommendation models either give only one recommendation (however,
there may be a variety of drug combination options in practice) or cannot provide the confidence level of the
recommended result. To fill these gaps, a Drug Recommendation model based on Message Propagation
neural network (denoted as DRMP) is proposed in this paper. Then, the Drug-Drug Interaction (DDI)
knowledge is introduced into the proposed model to reduce the DDI rate in recommended drugs. Finally, the
proposed model is extended to Bayesian Neural Network (BNN) to realize multiple recommendations and
give the confidence of each recommendation result, so as to provide richer information to help doctors make
decisions. Experimental results on public data sets show that the proposed model is superior to the best
existing models. [1].
• This study builds a drug recommendation system for patients who have diabetes. One method applied in a
recommendation is Case Base Reasoning (CBR). The CBR method calculates based on the similarity of the
old case to the new issue. In calculating the similarity of the problems using the Nearest Neighbors
algorithm. Although there have been many studies on drug recommendations, no studies have added risk
factor variables. Risk factors can influence results more accurately. The results of this study are in the form
of a prescription drug recommendation system for diabetic patients based on the type of diabetes they suffer
from. They are testing the accuracy of this system using the Confusion Matrix with an accuracy of 80.60\%.
Suggestions for further research are to add several types of similarity. It is necessary to research different
kinds of diseases. It is expected to produce an integrated system that can provide drug recommendations for
various conditions. [2].
• In this paper, we design and implement a drug recommender system framework that applies sentiment
analysis technologies on drug reviews. The objective of this research is to build a decision-making support
platform to help patients to achieve more significant choices in drug selection. Firstly, we propose a
sentimental measurement approach to drug reviews and generate ratings on drugs. Secondly, we take how
much the drug reviews are useful to users, patient's conditions, and dictionary sentiment polarity of drug
reviews into consideration. Then, we fuse those factors into the recommendation system to list appropriate
medications. Experiments have been carried out using Decision Tree, K-Nearest Neighbors, and Linear
Support Vector Classifier algorithm in rating generation and Hybrid model in recommendation based on the
given open dataset. The analysis is carried out to tune the parameters for each algorithm in order to achieve
greater performance. Finally, Linear Support Vector Classifier is selected for rating generation to obtain a
good trade-off among model accuracy, model efficiency, and model scalability. [3].
• This research focuses on prescriptions of GuiZhi Decoction, which is from Zhongjing Zhang's Treatise on
Cold Pathogenic Diseases. First, we constructed the database for prescriptions of GuiZhi Decoction based
on the prescriptions' names, the concept extraction and formal expression. Then we generate the formal
context and optimized it by use of the mathematical theory of formal concept analysis and the relationship
between prescriptions and herbs in database. In the formal context, the prescriptions are objects and the
herbs are attributes. Next the structural partial-ordered attribute diagram is generated, so as to visualize the
knowledge of prescriptions of GuiZhi Decoction. Finally, through the analysis of structural partial-ordered
attribute diagram, more comprehensive, objective and multileveled knowledge are discovered. [4].

MOTIVATION
The motivation behind developing a Machine Learning-based Doctor Receipt Analyzer system using OCR
technology stems from the desire to address significant challenges in the current healthcare administration
landscape. Manual processing of doctor receipts is time-consuming and prone to errors. The motivation is to
enhance the efficiency of healthcare administration by automating the extraction of crucial information from
receipts, reducing the burden on healthcare professionals. Handwritten or varied receipt formats can pose
challenges in accurate data extraction.

IJIRMPS230543 Website: www.ijirmps.org Email: editor@ijirmps.org 2

 Volume 12 Issue 2 @ 2024 IJIRMPS | ISSN: 2349-7300

SYSTEM ARCHITECTURE

Fig -1: System Architecture Diagram

DETAILS OF MODULE

• User Interface Module: This module provides a user-friendly web interface for healthcare professionals to
interact with the system. It includes features for user authentication, receipt uploads, and access to system
functionalities.
• Image Preprocessing Module: Responsible for optimizing receipt images before OCR processing. This
module ensures that images are standardized, enhancing OCR accuracy.
• Text Extraction and Parsing Module: Processes the OCR-converted text, extracting relevant information
such as patient names, dates, services provided, and costs. This module employs machine learning
algorithms for pattern recognition and data extraction.

APPLICATION
• Patient Record Integration: Integrates receipt data into patient records, creating a comprehensive view of
financial transactions associated with patient care.
• Research and Audit Support: Facilitates research activities and audit processes by providing accurate and
organized receipt data for analysis.
• Research and Audit Support: Facilitates research activities and audit processes by providing accurate and
organized receipt data for analysis.
• Enhanced Financial Transparency: Provides a transparent view of financial transactions, contributing to
improved financial transparency within healthcare institutions.

ALGORITHM
1. Data Input and Integration: The system integrates with electronic health records (EHRs) to retrieve
comprehensive patient information, including previous diagnoses and prescriptions.
2. Natural Language Processing (NLP): NLP algorithms process and analyze free-text inputs, extracting
relevant information from patient descriptions of symptoms and medical history.
3. Data Preprocessing: Preprocessing techniques are applied to clean and structure the data, including text
normalization, tokenization, and removal of stop words and irrelevant information.
4. Machine Learning for Symptom Analysis: Machine learning models analyze the preprocessed symptom
data, taking into account the patient's medical history, to generate a list of potential diagnoses and associated
probabilities.
5. Drug Database Query: The system queries a comprehensive drug database to retrieve information on
available medications, dosages, side effects, and potential interactions.

IJIRMPS230543 Website: www.ijirmps.org Email: editor@ijirmps.org 3

 Volume 12 Issue 2 @ 2024 IJIRMPS | ISSN: 2349-7300

6. Personalized Drug Recommendations: The machine learning algorithms consider factors such as the
patient's medical history, allergies, and potential drug interactions to generate personalized medication
recommendations based on the diagnosed condition.
7. Clinical Guidelines Integration: The system aligns medication recommendations with established clinical
guidelines and best practices to ensure that prescribed medications meet medical standards.
8. Prescription Analysis: Concurrently, the system evaluates doctor prescriptions in real-time, comparing
them to clinical guidelines, identifying potential drug interactions, and optimizing the medication choices
for individual patients.

RESULTS

IJIRMPS230543 Website: www.ijirmps.org Email: editor@ijirmps.org 4

Volume 12 Issue 2 @ 2024 IJIRMPS | ISSN: 2349-7300

CONCLUSION
In conclusion, the Machine Learning-based Doctor Receipt Analyzer system harnessing OCR technology
marks a significant leap forward in healthcare administration. This innovative solution has successfully
automated and streamlined the intricate process of processing doctor receipts, transcending the limitations
posed by diverse formats and handwritten entries. By seamlessly integrating machine learning algorithms,
the system not only enhances the accuracy and efficiency of data extraction but also promises to
revolutionize financial management and decision-making within healthcare institutions. Its adaptability,
commitment to data security, and potential for future advancements position it as a pivotal tool in the
evolving landscape of healthcare administration. As we embrace this cutting-edge technology, we anticipate
a future where administrative workflows are further optimized, allowing healthcare professionals to allocate
more time and resources to the core of their practice – delivering exceptional patient care.

REFERENCES:
[1] Telemedicine, https:/ /www. mohfw. gov.in /pdf/ Telemedicine.pdf
[2] Wittich CM, Burkle CM, Lanier WL. Medication errors: an overview for clinicians. Mayo Clin Proc.
2014 Aug;89(8):1116-25.
[3] CHEN, M. R., & WANG, H. F. (2013). The reason and prevention of hospital medication errors.
Practical Journal of Clinical Medicine,
[4] Drug Review Dataset, https:// archive. ics. uci. edu/ ml/ datasets/Drug% 2B Review% 2B Dataset%
2B%2528 Drugs.com% 2529#.
[5] Fox, Susannah, and Maeve Duggan. ”Health online 2013. 2013.”
[6] Bartlett JG, Dowell SF, Mandell LA, File TM Jr, Musher DM, Fine MJ. Practice guidelines for the
management of community-acquired pneumonia in adults. Infectious Diseases Society of America.
Clin Infect Dis. 2000 Aug;31(2):347-82. doi: 10.1086/313954. Epub 2000 Sep 7. PMID: 10987697;
PMCID: PMC7109923.
[7] Fox, Susannah & Duggan, Maeve. (2012). Health Online 2013. Pew Research Internet Project Report.
[8] T. N. Tekade and M. Emmanuel, ”Probabilistic aspect mining approach for interpretation and
evaluation of drug reviews,” 2016 International Conference on Signal Processing, Communication,

IJIRMPS230543 Website: www.ijirmps.org Email: editor@ijirmps.org 5

Volume 12 Issue 2 @ 2024 IJIRMPS | ISSN: 2349-7300

Power and Embedded System (SCOPES), Paralakhemundi, 2016, pp. 1471-1476, doi:
10.1109/SCOPES.2016.7955684.
[9] Doulaverakis, C., Nikolaidis, G., Kleontas, A. et al. GalenOWL: Ontology-based drug
recommendations discovery. J Biomed Semant 3, 14 (2012).
[10] Leilei Sun, Chuanren Liu, Chonghui Guo, Hui Xiong, and Yanming Xie. 2016. Data-driven
Automatic Treatment Regimen Development and Recommendation.

IJIRMPS230543 Website: www.ijirmps.org Email: editor@ijirmps.org 6

 Appendix B

PLGARISIM REPORT

91
 PLAGIARISM SCAN REPORT

Date 2024-05-03

19% 81% Words 484

Plagiarised Unique

Characters 4034

Content Checked For Plagiarism

n this inventive application, we present an AI based Specialist Receipt

Analyzer framework engaged by Optical Person Acknowledgment (OCR) innovation.
The framework addresses a groundbreaking jump in the handling of clinical receipts
via flawlessly coordinating high level OCR calculations. Fit for removing basic
data from filtered or shot receipts, the arrangement computerizes the tra-
ditionally manual assignment of translating transcribed or printed text, improving both
exactness and effectiveness. Custom fitted for easy mix into medical services administra-
tion, the framework brilliantly deciphers fundamental subtleties like patient names, dates,
administrations gave, and related costs. The cooperative energy of AI and OCR
smoothes out charging processes for medical care experts as well as commitments
to raise the general effectiveness of clinical documentation and monetary administration,
denoting a huge progression in medical care innovation

Presenting a state of the art arrangement, our AI based Specialist Receipt An-
alyzer framework saddles the force of OCR innovation to reform the way clinical
receipts are handled. By utilizing complex Optical Person Acknowledgment
calculations, the framework proficiently extricates pivotal data from filtered or pho-
tographed receipts. This creative methodology not just computerizes the customarily
manual and tedious assignment of translating transcribed or printed text yet additionally
upgrades precision and productivity. Consistently incorporating into medical services administra-
tion, the framework shrewdly interprets patient names, dates, administrations gave, and
related costs. The collaboration of AI and OCR not just smoothes out the
charging and record-saving interaction for medical care experts yet in addition vows to
lift the general productivity of clinical documentation and monetary administration.
This earth shattering AI driven Specialist Receipt Analyzer means a
outlook change in medical care the executives. Via flawlessly combining the abilities of
OCR innovation with cutting edge calculations, our framework changes the handling
of clinical receipts. Its ability lies in its capacity to skillfully remove essential data
from different receipt designs, be they filtered or shot. Past robotizing
the generally work serious undertaking of interpreting written by hand or printed text, the
framework altogether helps precision and functional effectiveness. Custom fitted to crease
lessly incorporate into medical care organization work processes, it shrewdly unravels key
components, for example, patient names, dates, administrations delivered, and associat

1. Yongjian Ren; Yuliang Shi, A Drug Recommendation Model Based on Message

Propagation and DDI Gating Mechanism, IEEE Journal of Biomedical and Health
Informatics ( Volume: 26, Issue: 7, July 2022): Drug recommendation task based on

Page 1 of 2
the deep learning model has been widely studied and applied in the health care field
in recent years. However, the accuracy of drug recommendation models still needs
to be improved. In addition, the existing recommendation models either give only
one recommendation (however, there may be a variety of drug combination options
in practice) or can not provide the confidence level of the recommended result. To
fill these gaps, a Drug Recommendation model based on Message Propagation neural
network (denoted as DRMP) is proposed in this paper.

Matched Source

Similarity 4%
Title:A Drug Recommendation Model Based on Message Propagation …
WEBDrug recommendation task based on the deep learning model has been widely studied and applied in the health care
field in recent years. However, the accuracy of A Drug Recommendation Model Based on Message Propagation and DDI
Gating Mechanism …
Check

Similarity 17%
Title:A Drug Recommendation Model Based on Message ...
by Y Ren · 2022 · Cited by 14 — However, the accuracy of drug recommendation models still needs to be improved. In
addition, the existing recommendation models either give only ...
https://pubmed.ncbi.nlm.nih.gov/35196249/

Similarity 4%
Title:ieeexplore.ieee.org › abstract › documentA Drug Recommendation Model Based on Message Propagation and ...
To fill these gaps, a Drug Recommendation model based on Message Propagation neural network (denoted as DRMP) is
proposed in this paper. Then, the Drug-Drug Interaction (DDI) knowledge is introduced into the proposed model to reduce
the DDI rate in recommended drugs.
https://ieeexplore.ieee.org/abstract/document/9720200/

Page 2 of 2