Breast cancer classification using deep learning

Project Title

Breast cancer classification using deep learning

Project Area of Specialization Artificial IntelligenceProject Summary

               Through the years, several CAD systems have been developed to help doctors and radiologists in finding the signs of cancer in screening mammogram tests. Many modules are designed to identify normal x-rays to relieve time for radiologists to focus on abnormal x-rays. But this module will spare the hours-long effort of radiologists in diagnosing cancerous cells in mammograms by using the image processing techniques and applying deep learning algorithms to classify the difference between benign, malignant tumors and other biological reasons from the x-rays. This module will also help in increasing accuracy of the test as we all know that while human detection, human error problem also rises and the accuracy matters. Also one of the key features of this module is that it will help in detecting cancer even before symptoms rise.

Project Objectives

The basic purpose of the project is to provide facility to the doctors and the radiologists in diagnosing breast cancer in radio images. Usually radiologists have to examine radio images for hours to find some abnormality in the tests. Even after diagnosing some abnormality one couldn't surely tell either its benign or malignant tumor.

            So this project take control and clears out with almost above 90% accuracy about benign and malignant with the help of deep learning algorithm. Objectives also include:

• • To identify benign and malignant tumor using x-rays.
  • Providing ease to the doctor and radiologist.

Time efficient for patients.

Project Implementation Method

The project implementation method involves the SVM and Decision tree algorithm in MATLAB using mobilenet v2 and Inception methods. Basically we will classify the dataset either benign or malignant, for that purpose we will run the dataset in matlab under Decision tree or Support vector machine algiorthms and will train the module using movile net v2 methods and more.

Benefits of the Project

In these days lot of women face breast cancer problem. Due to lack of awareness and post identification, this cancer will major reason of women deaths. CAD is giving the solution to diagnose this problem with accurate results but in this modern era of Artificial intelligence, it will be easy and reliable to diagnose this disease without any pain or wasting lot of money on tests. Classification with deep learning give a new way to diagnose the cancerous tumor with more accurate results and reduce the burden of radiologists in their relevant field. It help radiologist a lot to make a perfect decision in short time by analyzing the tumor. This system will give the better results, compared with previous methods like mammography, biopsy or FNA etc. The use of deep learning and neural networks and image processing helps to give accurate results in short time as compared with commercial hardware. So this system will give better results of tests in relevant field.

Technical Details of Final Deliverable

The final details will include:

• a fully trained model to identify breast cancer with accracy of 95% and above.
• a prototype for the model.
• FYP docmentation including 8 chapters and complete details related to the project.

Final Deliverable of the Project Software SystemCore Industry MedicalOther Industries Health Core Technology Artificial Intelligence(AI)Other Technologies Big DataSustainable Development Goals Good Health and Well-Being for People, Partnerships to achieve the GoalRequired Resources

Item Name	Type	No. of Units	Per Unit Cost (in Rs)	Total (in Rs)
			Total in (Rs)	74000
GeForce GTX 1660 SUPER Twin Fan ZT-T16620F-10L Graphics Card	Equipment	1	65000	65000
Stationary	Miscellaneous	1	9000	9000