bolimath / Breast-Cancer-Detection-Using-CNN-on-Ultrasound-Images

This Project is a 7 Layer CNN Model consisting of 3 Convolution layers each followed by a Max Pooling Layer and Fully Connected layer on Breast Ultrasound Images to classify them as Benign, Malignant and Normal stages.

Project Name : Breast-Cancer-Detection-Using-CNN-on-Ultrasound-Images

This Project is a 7 Layer CNN Model consisting of 3 Convolution layers each followed by a Max Pooling Layer and Fully Connected layer on Breast Ultrasound Images to classify them as Benign, Malignant and Normal stages.

• For each Convolution layer The Filter used is of size 3X3.

• We have used Relu as Activation function because it is non-linear in nature and due to its simplicity and less complexity.

• For every layer Stride(s) =1 and Padding(p) =0

• We have divided the dataset in 4:1 ratio for Model training and validation.

• For every Convolution layer result in CNN we use the formula

Project Implementation Video :

Implemetation_Video.mp4

Model Summary

Loss and Accuracy

Final Model Testing Prediction Result

_{^{Priyansh Sharma}}