Stanley college of engineering and technology for women

Mini project
on
Heart Disease Analysis and Prediction

INTERNAL GUIDE: : Dr K Ramakrishna, Professor of IT

Presented by:
Miss. Aanchal gupta(160617737001)
Miss .G.Preethi(160617737014)
Miss . Yerrolla Prasanna(160617737055)
 Contents
• Abstract
• Introduction
• Motivation
• Literature survey
• Existing Models &Comparison
• Problem Statement
• Proposed System
• Algorithm
• Flowchart
• Development Platform
• Software requirements
• Hardware Requirements
• Framework
• Result Analysis
• Output
• Conclusion
• Future scope
• References

stanley college of engineering technology for women IT-V SEM 2

 Abstract
Heart is the most important organ of a human body as it not only circulates oxygen and other vital nutrients
through blood to different parts of the body and helping in the metabolic activities but also removes metabolic
wastes. Thus, even a minor problem can affect the whole organism. Hence, researchers are diverting a lot of data
analysis work for assisting the doctors to predict the heart problems. We develop a heart disease prediction
system that can assist medical professionals to provide effective treatment to patients. The proposed Heart
Disease Prediction Analysis (HDPA) is used to predict the severity level of heart diseases, fast and effectively.
HDPA is designed and implemented to evaluate the performance of ANN and other prediction algorithms. We
took 14 important features from clinical dataset and developed a model with an Artificial Neural Network(ANN)
algorithm for classifying heart disease. The principle goal of this project is to transform information into
valuable data that can empower medical professionals to decide the successful treatment decision. This model
predicts True Positive Rate and False-Negative Rate of the patients based on given clinical data and shown
better results when compared with SVM, KNN, Naive Bayes etc.

stanley college of engineering technology for women IT-V SEM 3

 Introduction
The heart is a most significant muscularorgan in humans, which pumps blood through the blood vessels of the
circulatory system. Human life is dependent on the proper functioning of heart. Improper functioning of heart
will influence other parts of human body like brain, kidney etc. If the blood circulation in body is inefficient, it
affects both heart and brain. Generally blood arrest in heart is called as attack and blood arrest in brain is called
as stroke. Human life is absolutely reliant on the efficient working of the heart and brain.
Cardiovascular disease:
Cardiovascular heart disease is one of the principal reasons of death for both men and women. The term heart
disease relates to a number of medical conditions related to heart which define the irregular health conditions
that directly stimulate the heart and all its parts. Different types of heart related cardiovascular diseases along
with description are given below.

stanley college of engineering technology for women IT-V SEM 4

Heart-related cardiovascular diseases Description
Acute coronary syndromes Blood-supply to the heart muscle is swiftly
obstructed
Arrhythmia Atypical heart rhythm
Cardiomyopathy Heart muscle disease
Congenital heart disease Heart disfigurements that are present at birth
Coronary heart disease Arteries supplying blood to heart muscle becomes
obstructed
Heart failure Heart is not propelling ample blood
Inflammatory heart disease Tenderness of the heart muscle and/or the tissue
Ischaemic heart disease Plaque builds up inside the coronary arteries
Rheumatic heart disease Rheumatic fever
Valvular disease Disease of the valves

stanley college of engineering technology for women IT-V SEM 5

• An Artificial intelligence system has been developed using Artificial Neural Network(ANN) algorithms to diagnosis heart
disease from phonocardiogram(PCG) signals. The diagnosis of heart disease in most cases depends on a complex
combination of clinical and pathological data. Because of the complexity, there exists a significant amount of interest
among clinical professionals and researchers regarding the efficient and accurate prediction of heart disease. We develop a
user-friendly Heart Disease Predict System (HDPS). The HDPS system will be consisted of multiple features, including
input clinical data section, Receiver Operating Characteristic(ROC) curve display section, and prediction performance
display section(execute time, accuracy, sensitivity, specificity, and predict result).
• Our approaches are effective in predicting the heart disease of a patient. It include three steps.
• Firstly, we take 14 important clinical features, i.e., age, sex, chest pain type, restbps(resting blood pressure), chol(serum
cholesterol level), fbs(fasting blood sugar),restecg(resting electrocardiography), Thalach (maximum heart rate achieved),
exang(exercise induced angina), old peak(ST depression induced by exercise relative to rest), slope, ca(number of major
vessels colored by fluroscopy) ,num(diagnosis of heart disease ) and thal.
• Secondly, we develop an Artificial Neural Network(ANN) algorithm for classifying heart disease based on these clinical
features. The accuracy of prediction is near 80%.
• Finally, we develop a user-friendly heart disease prediction system(HDPS).

stanley college of engineering technology for women IT-V SEM 6

 Motivation
• The main objective of this project is to develop a system for diagnosis of Heart disease using Artificial Neural
Network modelling technique. This system can discover and extract hidden knowledge associated with heart
disease from historical heart disease database. It assists healthcare practitioners to make intelligent clinical
decisions which traditional decision support systems cannot. Here, system can classify the patients based on
risk level using Artificial Neural Network technique into two classes as Low, High. So, these diagnosis helps
healthcare practitioners in taking positive and effective decision in following condition:
1) Patient at high risk requires lowering treatment and special attention.
2) Patient at low risk and no risk may be encouraged by their doctor to follow Health
Recommendation for prevention of heart disease.
• So, the proposed system significantly acts as decision support system for doctors in taking clinical
decisions

stanley college of engineering technology for women IT-V SEM 7

 Literature Survey
• Shadab Adam Pattekari et al. [35] developed a prototype of Heart Disease Prediction System using Naive
Bayes, Decision trees and neural networks. It is implemented in a web application. In this system, user
answers some predefined questions. Then it retrieves hidden data from the stored database and compares the
user’s values with trained dataset. The system discovers and extracts hidden knowledge associated with heart
diseases from a historical heart disease database. It can answer the complex queries for diagnosing a disease.
A set of 15 attributes was selected and then Naive Bayes classification method was applied to find out the
chances of heart disease.
• Noura Ajam [38] has studied that artificial neural networks(ANN) show significant results in heart disease
diagnosis. The architecture of a neural network is formed by the number of processing units (neurons) and
connections between them. A subgroup of processing elements is called layer. The number of neurons and
the layers depends upon the complexity of the system. Artificial neural network is widely used in medical
diagnosis and health care applications because of it’s high predictive power as classifier, fault tolerance and
learning from environment. Artificial neural network is unsupervised learning type provided only with inputs
associated with unknown targets. It is self organized. The dataset used here is obtained from Cleveland
dataset which consists of 14attributes and 303 instances. Artificial neural network is trained using back
propagation learning algorithm on the data. Input and target samples are divided as 60% training set, 20%
validation set and 20% test set. The activation function used is tangent sigmoid for hidden layers and linear
transfer function for output layer. Mean square error (MSE) is calculated which is equal to 0.1071 and the
classification accuracy for heart disease is 88%.

stanley college of engineering technology for women IT-V SEM 8

 Existing Models &Comparison
• Mostly the work related to the prediction and figuring out the heart problem, many data driven techniques has been
used in past and the work inclines towards the classification problem. This is a process used to tune a model and
then predict the class for whether the patient is suffering from any heart related problem or not. To talk about the
intelligent methods in the medical sector, a vast number of related works has been performed .
• The six algorithms considered for the prediction of heart disease patient based on the accuracy, sensitivity and
specificity.

Algorithm Accuracy Sensitivity Specificity

ANN 85.30% 83.75% 75.73%
Naïve Bayes 81.14% 61.03% 70.11%
Decision Support 79.05% 83.12% 74.26%
SVM 85.97% 90.10% 77.20%
KNN 84.12% 56.87% 71.21%
RIPPER 81.08% 86.25% 75.82%

stanley college of engineering technology for women IT-V SEM 9

• The Naive Bayes or Bayes’ Rule is the basis for many machine learning and data mining methods. The
rules(algorithm) is used to create models with predictive capabilities .It provides new ways of exploring and
understanding data. I learns from the “evidence” by calculating the correlation between the largest and other
variables.
• Three popular data mining algorithms (support vector machine, artificial neural network and decision tree)
were employed to develop a prediction model using 502 cases for better prediction of heart problems . SVM
has the best accuracy but it is more sensitive when compared to ANN.
• Algorithmic complexity and memory requirements of SVM are very high. You need a lot of memory since
you have to store all the support vectors in the memory and this number grows abruptly with the training
dataset size.
• In SVM sensitivity is more when compared to ANN which effect the prediction of the result .To overcome
these sensitivity problem and also training data set fast we are using ANN and training our model to predict
the result.

stanley college of engineering technology for women IT-V SEM 10

 Problem Statement
• Many hospital information systems are designed to support patient billing, inventory management and
generation of simple statistics. Some hospitals use decision support systems, but they are largely limited. They
can answer simple queries like “What is the average age of patients who have heart disease?”, “How many
surgeries had resulted in hospital stays longer than 10 days?”, “Identify the female patients who are single,
above 30 years old, and who have been treated for cancer.” However, they cannot answer complex queries
like “Identify the important preoperative predictors that increase the length of hospital stay”, “Given patient
records on cancer, should treatment include chemotherapy alone, radiation alone, or both chemotherapy and
radiation?”, and “Given patient records, predict the probability of patients getting a heart disease.” Clinical
decisions are often made based on doctors’ intuition and experience rather than on the knowledge rich data
hidden in the database. This practice leads to unwanted biases, errors and excessive medical costs which
affects the quality of service provided to patients. So we proposed that integration of clinical decision support
with computer-based patient records could reduce medical errors, enhance patient safety, decrease unwanted
practice variation and improve patient outcome. This suggestion is promising as data modelling and analysis
tools e.g., neural network, have the potential to generate a knowledge-rich environment which can help to
significantly improve the quality of clinical decisions

stanley college of engineering technology for women IT-V SEM 11

 Proposed system

• An artificial neural network is a computational network model (neurons) that are massively interconnected,
operate in parallel. These elements are inspired by biological nervous system. The associations between
elements to a great extent decide the system work .Therapeutic Determination utilizing artificial neural
networks is at present an exceptionally dynamic research territory in prescription and it is trusted that it will
be all the more generally utilized as a part of biomedical frameworks.
A. Data Preparation
• The data preparation is the important part to determine the best variables that contribute to define the
optimal solution. This requires the detailed knowledge about the problem domain and the primary data’s. Here
the selection of this primary data is based on the doctor’s suggestions and the survey of the literatures.

stanley college of engineering technology for women IT-V SEM 12

Attributes Description Weights

Age for both male and female Age<30 0.1

30 to 50 0.4
50 to 70 0.7
>70 0.9
Family history Yes 0.5
No 0.1
Chest pain Type: typical angina 0.6
atypical angina 0.7
non-angina pain 0.8
asymptomatic 0.9
Smoking Never 0.1
Past 0.4
Current 0.7
Poor diet Yes 0.9
No 0.2
Trestbps(Blood pressure)in mm Normal(130/89) 0.1
Hg Low(200/169) 0.8
High(>200/169) 0.9
Blood sugar(Fbs) High(>120&<400) 0.6
Normal(>90&<120) 0.1
Low(<90) 0.4

Serum cholesterol in mg/dl Very high>200 0.9

High 160 to 200 0.8
Normal<160 0.1
Thalach(heart rate) High 0.9
Normal 0.1
Lowstanley college of engineering technology for women 0.9
IT-V SEM 13
Exang (Exercise) Never 0.7
Regular 0.1
High if age<30 0.1
High if age<50 0.8
Physical inactivity Yes 0.8
No 0.2
Obesity Yes 0.8
No 0.1
High Stress Yes 0.7
No 0.2
Poor hygiene Yes 0.6
No 0.2
Alcohol Yes 0.7
Intake No 0.2
Rest ecg (Resting electrocardiographic results) Having ST-T wave abnormality Showing probable or 0.6
define left ventricular hypertrophy
0.8

stanley college of engineering technology for women IT-V SEM 14

• The neural network is trained with the significant dataset or attributes as their inputs for the efficient heart
disease prediction. The prediction is based on the computed major weights given in the table for each
attribute. In ANN the processing element is called a layer.
• The first layer is known to be the input layer and the last layer is the output layer. Between the input and the
output layer the additional layers of units is called as hidden layer. A function can be performed by adjusting
the values of the weights between elements
Data Preprocessing
• The collected data were used to create a structured database system. The fields were identified, duplications
were extracted, missing values were filled and the data were coded according to attribute domain value. After
data cleaning the number of cases was reduced mainly due to unavailability of clinical result. Attribute
domain values are provided by practicing cardiologist.

stanley college of engineering technology for women IT-V SEM 15

 Algorithm
• Figure shows the structure of MLNN(Multi level neural network) with the layers were the significant inputs
to heart disease prediction have been used with their weights given in the table.In which 0.1 refers to the
lower risks and >0.1 refers to the higher risks.

stanley college of engineering technology for women IT-V SEM 16

• Here for the prediction of the heart disease the Multi-layer Neural Network with Back-Propagation is being used
as the training algorithm. This is done in a unidirectional way. In a feed forward neural network mechanism, the
input neurons of the first layer[X1, X2, X3….Xn] forward their output to the neurons of the second layer as [V11,
V21…] from Which hidden layer [Y1,Y2,…Ym] evaluates the weights [W11,W22,……] and gives the output
layer as [z1…]. By using this algorithm heart disease diagnosis is done with the input attributes and the weights
given in the table . The back propagation equation which is represented below is used for back propagation neural
network.

stanley college of engineering technology for women IT-V SEM 17

 FLOWCHART
TRAINING DIAGNOSIS
Features Patient(s) to
Start
selections be examined

ANN and
patient data DATABASE
collection
DATABASE building
diagnosis

preprocessing
Evaluated/predicted
result

ANN training and

verification New Y
patient

N
Verified
DATABASE End

stanley college of engineering technology for women IT-V SEM 18

 Development Platform

• Software Requirements
Programming language: Python (version:2.1.3)
Operating System:Windows-10
• Hardware Requirements
RAM:512MB
Hard Disk:20GB
Processor: Intel core
• Framework
Framework_name: Tensorflow
Framework_version: 1.5

stanley college of engineering technology for women IT-V SEM 19

 RESULT ANALYSIS
This proposed system classifies the heart disease data into 5 categories diseases with 97.5% accuracy
using backpropagation algorithm. In this paper, the prediction system gives the improved result with
highest accuracy of 98.58% for 20 neurons in hidden layer with same heart disease database. The
multi-layered decision making system embedded with improved algorithm proved to be effective by
dividing its training dataset on multiple subsets with 82.8% accuracy with running time of
5.97seconds. The prediction system in this paper gives higher accuracy of 93.39% for 5 neurons in
hidden layer with running time of 3.86seconds. The performance of the proposed system is compared
with other classification techniques. This shows that the proposed prediction system shows higher
performance

stanley college of engineering technology for women IT-V SEM 20

 Output

stanley college of engineering technology for women IT-V SEM 21

stanley college of engineering technology for women IT-V SEM 22
stanley college of engineering technology for women IT-V SEM 23
 Conclusion
• The principle target of this examination is to assess the utilization of artificial neural network with Multi layer
neural network in analysis of heart diseases. The performance of back propagation neural network with the
prediction accuracy is satisfactory but to increase the level of accuracy and to know the type of heart disease
the CBR technique is integrated with ANN. The data has been collected in relation to heart patients and the
analysis has been done on the probability of heart attack to a person with certain medical criteria. The raw
data has been processed and processed information is used in automated expert system to guide in diagnosing
coronary heart disease which results in finding new and accurate rule set that relate patient data to the heart
disease risk level and can provide more useful and exact information. It is proven from the results that the
proposed method effectively predicts the heart disease through the 14 attributes when compared to the other
approaches. This work can be extended to predict and analyze the level of the disease by considering more
attributes. So the accuracy of 97% is adapted by using the old patient records. CBR is not only used to
increase the accuracy but also to predict the type of heart disease.

stanley college of engineering technology for women IT-V SEM 24

 FUTURE SCOPE

• In future an intelligent system may be developed that can lead to selection of proper treatment
methods for a patient diagnosed with heart disease. A lot of work has been done already in making
models that can predict whether a patient is likely to develop heart disease or not. There are several
treatment methods for a patient once diagnosed with a particular form of heart disease. Data
mining can be of very good help in deciding the line of treatment to be followed by extracting
knowledge from such suitable databases

stanley college of engineering technology for women IT-V SEM 25

 References
• S. Vijaya ani, S. Sudha, “Comparative Analysis of Classification Function Techniques for Heart
Disease”, International Journal of Innovative Research in Computer and Communication Engineering,
Vol. 1, Issue 3, May 2013.
• M.Durairaj, K. Meena, - A Hybrid Prediction System Using Rough Sets and Artificial Neural
Networks‖, International Journal of Innovative Technology & Creative Engineering (ISSN: 2045-8711)
VOL.1 NO.7 JULY 2011.
• Anbarasi M., Anupriya E., Iyengar N.C.S.N., Enhanced prediction of heart disease with feature subset
selection using genetic algorithm, International Journal of Engineering Science and Technology 2(10)
(2010), 5370-5376.
• Awang M.K., Siraj F., Utilization of an Artificial Neural Network in the Prediction of Heart Disease,
International Journal of Bioscience and Bio-Technology 5(4) (2013), 159-166.

stanley college of engineering technology for women IT-V SEM 26