MRANet: A Modified Residual Attention Networks for Lung and Colon Cancer Classification

TL;DR

This paper introduces MRANet, a modified residual attention network that achieves high accuracy in classifying lung and colon cancers from histopathological images, outperforming existing models.

Contribution

The study presents a novel deep learning architecture tailored for cancer classification, demonstrating superior accuracy on a large dataset compared to prior methods.

Findings

Achieved up to 99.30% accuracy for two-class classification.

Outperformed other state-of-the-art architectures.

Validated on a dataset of 25,000 images.

Abstract

Lung and colon cancers are predominant contributors to cancer mortality. Early and accurate diagnosis is crucial for effective treatment. By utilizing imaging technology in different image detection, learning models have shown promise in automating cancer classification from histopathological images. This includes the histopathological diagnosis, an important factor in cancer type identification. This research focuses on creating a high-efficiency deep-learning model for identifying lung and colon cancer from histopathological images. We proposed a novel approach based on a modified residual attention network architecture. The model was trained on a dataset of 25,000 high-resolution histopathological images across several classes. Our proposed model achieved an exceptional accuracy of 99.30%, 96.63%, and 97.56% for two, three, and five classes, respectively; those are outperforming other state-of-the-art architectures. This study presents a highly accurate deep learning model for lung and colon cancer classification. The superior performance of our proposed model addresses a critical need in medical AI applications.