Breast Cancer Detection: CNN Model Training

Project:
Training a CNN Model on Mammography Dataset for Breast Cancer Detection

Project Objective
Build a Convolutional Neural Network (CNN) for classifying breast cancer images.

Use CBIS-DDSM, MIAS, and INbreast datasets to train the model.

Evaluate the model’s performance using the following metrics:

Accuracy: Achieve up to 95%.
Recall/Sensitivity: Achieve up to 95%.
Specificity: Achieve up to 95%.
Precision.
F1 Score.

Note:
Implement the project using Python with GPU Tensorflow on native-Windows (local environment lab)
Note: 1. Preparing the Environment
Steps:
Install libraries:
Setup the work environment: Use Jupyter Notebook

2. Collect data
Steps:
Get a database like CBIS-DDSM, MIAS and INbreast.
Upload data to the working folder.
Organise files by categories:
benign/: images of benign tumors.
malignant/: images of malignant tumors.
normal/: normal images.

3. Read and explore the data
4. Preprocess the data (Data Preprocessing)
Steps:
Image scaling:
Normalize the image size to 224x224.
Data Augmentation:
Add operations like rotation and flip to improve diversity.
Image normalization:
Split pixel values into the range [0, 1].

5. Data segmentation
6. Model Building
7. Model training
8. Model evaluation
9. Results analysis
10. Model deployment