Bladder-Tissue-Classification-using-CNN

Implemented and trained convolutional neural network(CNN) from scratch using matlab and keras with tensorflow back-end to classify bladder pathology image regions into soft tissue, muscle, and tumor. Applied patch based classification as size of pathology images is very large. Evaluated CNN and CNN + Support Vector Machine(SVM)

The implemetation has MATLAB and Python version. MATLAB implemetation is tested in version 2016b, while python version is tested in python3.6 and

Libraries for python code

The code has been tested with the following configuration

• keras == 2.1.2
• python == 3.6
• scipy == 0.19.0
• sckit-learn == 0.18.1
• tensorflow == 1.4.1
• tensorflow-gpu == 1.4.1

Libraries for matlab code

The code has been tested with the following configuration

• matlab2016

• matconvnet24

How to run it

Once all the libraries above have been installed, download the files and run them.

sample output

Dice index evaluation

ROC Curve

Confusion matrix