BI-RADS CLASSIFICATION OF BREAST CANCER

MOMMOGRAPHIES

by
Cengizhan Şahin

Faculty of Engineering
Department of Computer Engineering
Istanbul, 2021

Yeditepe University
2021
 BI-RADS CLASSIFICATION OF BREAST CANCER
MOMMOGRAPHIES

APPROVED BY:

Prof. Dr. Name Surname (Supervisor) Emin Erkan Korkmaz

DATE OF APPROVAL:
 ACKNOWLEDGEMENTS

It is with immense gratitude that I acknowledge the support and help of my Professor Emin Erkan
Korkmaz. Pursuing my thesis under her supervision has been an experience which broadens the mind
and presents an unlimited source of learning.

Finally, I would like to thank my family for their endless love and support, which makes everything more
beautiful.
 ABSTRACT

BI-RADS CLASSIFICATION OF BREAST CANCER MAMMOGRAPHIES

Breast cancer is the most common cancer diagnosed in women, accounting for more than 1 in
10 new cancer diagnoses each year. It is the second most common cause of death from cancer among
women in the world. Breast cancer is the uncontrolled growth of abnormal cells in one of several areas
of the breast. Mammography is one of the most reliable tools for screening and a key method for the
early detection of breast cancer. It is one of the most suitable techniques to detect breast cancer. In
order to prevent deaths due to breast cancer and reduce the overwhelm on radiologist, deep learning
algorithms used to detect breast cancer and its type. In light of this, this project presents a deep neural
network (DNN) based model as a tool for people to use. Mammograms are classified into BI-RADS
categories as 0,1,2,3,4A,4B,4C and 5 but in the dataset, we used has labelled as BI-RADS 0, BI-RADS 1-2,
BI-RADS4-5. This project proposes 2 pipelines for achieve the maximum accuracy possible over this data
set. Several CNN’s and methods were analyzed depend on influencing the accuracy.
Table of Contents
ACKNOWLEDGEMENTS................................................................................................................................3
ABSTRACT....................................................................................................................................................4
LIST OF FIGURES..........................................................................................................................................7
LIST OF TABLES............................................................................................................................................8
LIST OF SYMBOLS/ABBREVIATIONS.............................................................................................................9
1. INTRODUCTION.................................................................................................................................10
1.1. BREAST CANCER AND MAMMOGRAMS.....................................................................................10
1.1.1. Breast Cancer.....................................................................................................................10
1.1.2. Mammograms....................................................................................................................10
1.1.3. Deep learning on mammograms........................................................................................11
1.2. Terms.........................................................................................................................................11
LIST OF FIGURES

Figure 3.1 Caption about first image....................................................................................10

Figure 3.2 Caption about multiple images. (a) First of multiple image, (b) Second of multiple

image, (c) Third of multiple image......................................................................................11

Figure 3.3 Growth curves of the wild type (AS102) and recombinant AS102-57..............11
LIST OF TABLES

Table 3.1 Sequence of the preliminary..................................................................................5

Table 3.2 Caption about table..............................................................................................10

LIST OF SYMBOLS/ABBREVIATIONS

A System matrix
B Input matrix
G Plant transfer matrix
G1 Plant disturbance matrix
R A closed and bounded region in E
δR Boundary of the region R
δRs s=1,…,p Complementary regular subsurface of the boundary δR, s being a positive scalar
T Sampling time
U Control Input
xc Controller state vector
αc, αe Controller and estimator characteristics polynomials
Г Discrete plant control input matrix
Г1 Discrete plant noise input matrix
λ Plant delay time or transportation lag

ADD Average detection delay

ASN Average sample number
i.i.d. independently and identically distributed

JACS Journal of American Chemical Society

1. INTRODUCTION

1.1. BREAST CANCER AND MAMMOGRAMS

1.1.1. Breast Cancer

Breast cancer is a type of cancer that forms in the cells of the breasts. It can occur in both men and
women, but it's far more common in women. Breast cancer can start in different parts of the breast such
as ducts, lobules, or tissue. The most common type of breast cancer is ductal carcinoma, which begins in
the cells of the ducts. Another type of breast cancer is lobular carcinoma, which begins in the lobules.

Breast cancer can be detected early by mammography screening tests. The American Cancer Society
recommends that women aged 40 to 44 should have the choice to start annual breast cancer screening
with mammograms if they wish to do so. Women aged 45 to 54 should get mammograms every year.
Women aged 55 and older should switch to mammograms every 2 years or can continue.

1.1.2. Mammograms

A mammogram is an X-ray image of the breast. It can be used either for breast cancer screening or
for diagnostic purposes, such as to investigate symptoms or unusual findings on another imaging test.
During a mammogram, your breasts are compressed between two firm surfaces to spread out the
breast tissue. Then an X-ray captures black-and-white images that are displayed on a computer screen
and examined for signs of cancer. Mammograms play a key role in breast cancer screening. They can
detect breast cancer before it causes signs and symptoms.

Figure 1.1. Mammogram

1.1.3. Deep learning on mammograms

Deep learning is a family of machine learning techniques that has spurred much interest in its
application to medical imaging problems. One such application is improving breast cancer detection on
screening mammograms. A deep learning algorithm can accurately detect breast cancer on screening
mammograms using an “end-to-end” training approach that efficiently leverages training datasets with
either complete clinical annotation or only the cancer status (label) of the whole image. This approach
eliminates the reliance on rarely available lesion annotations and can attain high accuracy on
heterogeneous mammography platforms. These findings hold tremendous promise for improving clinical
tools to reduce false positive and false negative screening mammography results.

1.2. TERMS

 Machine learning is a branch of artificial intelligence and computer science which focuses on the
use of data and algorithms to imitate the way that humans learn, gradually improving its
accuracy. Machine learning algorithms build a model based on sample data, known as training
data, in order to make predictions or decisions without being explicitly programmed.

 False positive and false negative are an error that occurs when machine learning algorithms
binary predictions as 1 or 0 are false. There are true positive and true negative outcomes that
opposite of these errors. True positive and true negatives are correct prediction. Positive as 1
and negative as 0. When algorithm predict a positive data as negative, that will become false
negative. There is confusion matrix to visualize the errors.

Figure 1.2 Confusion Matrix

 (2) What are false positive and false negative in information security?.
https://gatefy.com/blog/what-are-false-positive-and-false-negative/ Accessed 3/30/2023.
(3) Difference Between False Positive and False Negative.
http://www.differencebetween.net/science/difference-between-false-positive-and-false-
negative/ Accessed 3/30/2023.
(4) Classification: True vs. False and Positive vs. Negative.
https://developers.google.com/machine-learning/crash-course/classification/true-false-
positive-negative Accessed 3/30/2023.

Source: Conversation with Bing, 3/30/2023(1) Machine learning - Wikipedia.

https://en.wikipedia.org/wiki/Machine_learning Accessed 3/30/2023.

(2) What is Machine Learning? | IBM. https://www.ibm.com/topics/machine-learning Accessed

3/30/2023.

(3) . https://bing.com/search?q=what+is+machine+learning Accessed 3/30/2023.

(4) Machine learning, explained | MIT Sloan. https://mitsloan.mit.edu/ideas-made-to-matter/machine-

learning-explained Accessed 3/30/2023.

Source: Conversation with Bing, 3/30/2023(1) Deep Learning to Improve Breast Cancer Detection on
Screening ... - Nature. https://www.nature.com/articles/s41598-019-48995-4 Accessed 3/30/2023.

(2) Applying Deep Learning Methods for Mammography Analysis and Breast ....
https://www.mdpi.com/2076-3417/13/7/4272 Accessed 3/30/2023.

(3) Deep Learning to Improve Breast Cancer Detection on Screening Mammography.

https://pubmed.ncbi.nlm.nih.gov/31467326/ Accessed 3/30/2023.

[1] What Is Breast Cancer? | CDC. https://www.cdc.gov/cancer/breast/basic_info/what-is-breast-

cancer.htm

[2] Breast cancer - Symptoms and causes - Mayo Clinic. https://www.mayoclinic.org/diseases-

conditions/breast-cancer/symptoms-causes/syc-20352470 Accessed 3/29/2023.

[3] Breast cancer in women - NHS. https://www.nhs.uk/conditions/breast-cancer/ Accessed 3/29/2023.

(4) Breast cancer - World Health Organization.

https://www.who.int/news-room/fact-sheets/detail/breast-cancer Accessed 3/29/2023.
(5) . https://bing.com/search?q=what+is+breast+cancer Accessed 3/29/2023.

(6) What Is Breast Cancer? | American Cancer Society.

https://www.cancer.org/cancer/breast-cancer/about/what-is-breast-cancer.html Accessed 3/29/2023.

Source: Conversation with Bing, 3/30/2023(1) Mammogram - Mayo Clinic.

https://www.mayoclinic.org/tests-procedures/mammogram/about/pac-20384806 Accessed 3/30/2023.

Meme kanseri ölüm oranı hakkında

Neden bu kadar ölüm oranı yüksek

Mamografi ve meme kanseri tespiti

Meme kanseri mamografisi hakkında bilgiler ve yöntem LMLO LCC tarzında

Yapay zeka nasıl çözüm getiriyo

Hangi yöntemler bu alanda kullanılıyor

Veriseti hakkında bilgi

In preprocessing stage Yolo v5 Instance Segmentation model was trained and used cropping
mammograms to reduce the noise. After that, the mirroring method was applied to each LCC and LMLO
mammograms and they all became in the same orientation as the RCC AND RMLO methods. In our 2
pipelines were proposed to get best possible accuracy over this data set. First pipeline creates a feature
vector of LCC, LMLO, RCC and RMLO by using pretrained CNN’s for feature extraction and the feature
vector is given as input to a ANN classifier which have 3 hidden layers. Second pipeline proposing
ensemble

Referenslar :

Source: Conversation with Bing, 3/29/2023(1) Breast Cancer Detection and Diagnosis Using
Mammographic Data .... https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6688437/ Accessed
3/29/2023.
(2) . https://bing.com/search?
q=breast+cancer+mammography+detection+academic+article+introduction Accessed 3/29/2023.

(3) Mammography - National Institute of Biomedical Imaging and Bioengineering.

https://www.nibib.nih.gov/science-education/science-topics/mammography Accessed 3/29/2023.

(4) Detection of Breast Cancer with Mammography: Effect of an Artificial ....

https://pubmed.ncbi.nlm.nih.gov/30457482/ Accessed 3/29/2023.

Source: Conversation with Bing, 3/29/2023(1) Breast Cancer - StatPearls - NCBI Bookshelf.
https://www.ncbi.nlm.nih.gov/books/NBK482286/ Accessed 3/29/2023.

(2) Breast Cancer - Harvard Health. https://www.health.harvard.edu/topics/breast-health-and-disease

Accessed 3/29/2023.

(3) Introduction to Breast Cancer | Yale Online. https://online.yale.edu/courses/introduction-breast-

cancer Accessed 3/29/2023.

(4) Breast cancer - World Health Organization.

https://www.who.int/news-room/fact-sheets/detail/breast-cancer Accessed 3/29/2023.

(5) (PDF) Breast cancer: Introduction - ResearchGate.

https://www.researchgate.net/publication/257458575_Breast_cancer_Introduction Accessed
3/29/2023.

(6) Breast Cancer: Introduction - Health Encyclopedia - University of ....

https://www.urmc.rochester.edu/encyclopedia/content.aspx?contenttypeid=85&contentid=p09506
Accessed 3/29/2023.