Breast cancer is the most prevalent cancer among women, and it is also the second greatest cause of cancer mortality among them. Because the etiology of breast cancer is unknown, there are currently no effective techniques to prevent or cure it. Breast cancer can be diagnosed and managed more effectively if caught early, which increases the chances of a full recovery. Deep Learning has been shown to be the most effective method for detecting breast cancer in its early stages, and it is still the dominant imaging modality for breast cancer screening and diagnosis. This technique can also identify various diseases and may indicate the type of cancer, such as normal, benign, or malignant. This thesis explains how to extract characteristics from a picture to locate cancer-affected areas, which is an important stage in the diagnosis and verification of breast cancer.
This Project is built for following purpose:-
- Early Detection
- Best Accuracy
- Segment the tumor part
- User Friendly
- OpenCv
- matplotliv
- Image
- numpy
- keras
- pandas
- os(for handle folder data)
- sklearn
- seaborn
Link- http://peipa.essex.ac.uk/pix/mias/
DDSM- https://data.mendeley.com/datasets/ywsbh3ndr8/1
- Make root directory name - mini-MIAS
- Inside mini-MIAS folder create two folder - all-mias and data
- Inside all-mias folder paste all dataset image given above link
- And in the data folder create two more folder - benign_cases and malignant_cases
- Then run the project and handle all image_directory carefully
Note:- This is my final year project in which first i am segment the images with the help of python library openCV after that building the deep learning model and then classify the external images whether there is malignant or benign.
- Email: harshit.shakya99@gmail.com
- Website: https://harshitshakya123.github.io/
- LinkedIn: https://www.linkedin.com/in/harshit-shakya123/
Thanks a lot
Project Creator
Harshit Shakya