A decentralized learning-based approach to classify colorectal cancer using Deep Learning Leveraging XAI

Abstract

Convolutional Neural Networks (CNN)-based automated approaches are vastly utilised to anticipate and diagnose cancer, saving time and reducing mistakes. Deep Learning (DL) CNN methods use a variety of probabilistic and statistical methodologies to make com puters understand and identify patterns in datasets based on previous experiences. We proposed an efficient federated learning based model to classify histopathological images for detecting colorectal cancer while providing high prediction accuracy and maintaining data privacy. Federated learning solves the problem of retaining privacy while utilizing vast and heterogeneous private datasets collected from numerous healthcare facilities. As the amount of patient data obtained for the process of machine learning is significantly responsible for the success of enhancing the accuracy of the system, the experiment was performed on a large dataset including cancerous and non-cancerous colorectal tissue im ages. FL can also mitigate costs resulting from traditional, centralized machine learning approaches. We have also applied the XAI method, a model-agnostic approach to acquire an explicit demonstration of the applied machine learning models. With XAI, we can visualize the super pixels of our colorectal tissue images through accepting and reject ing features. While applying various CNN models such as VGG16 & 19, InceptionV3, ResNet50, ResNeXt50, and comparing their precision, ResNeXt50 was established with the highest accuracy of 99.53%. Therefore, we have applied ResNeXt50 on FL that brings forth the accuracy of 96.045% and F1 Score is 0.96.