Tumor classification using Convolutional Neural Networks

Roberto Loi

;Monica Pibiri

2025-01-01

---

Abstract

Tumor-classification-using-Convolutional Neural Networks. This is a software project for tumor classification with Convolutional Neural Networks. This software repository contains code that can be used when using a Convolutional Neural Network for tumor classification. It can be adapted to different classification tasks that involve 2 output classes. The code is written in Matlab programming language and it is commented. It can be used and adapted to the user context. The files contain scripts that can be used to extract Normal and Tumor blocks from multiresolution blocked images derived from Whole Slide Images (WSI) of normal and tumor specimens. The blocks can be used to train a CNN. A script that can be used train a CNN through transfer learning is also provided. The present software is published and distributed through the electronic platform GitHub (www.GitHub.com) (url: https://github.com/rloi66/Tumor-classification-using-Convolutional-Neural-Networks). License: MIT (Copyright © 2025 Roberto Loi and Monica Pibiri)