Research on Cancer Detection Published by Researchers at University of Molise (MR_NET: A Method for Breast Cancer Detection and Localization from Histological Images Through Explainable Convolutional Neural Networks).

Researchers at the University of Molise have developed a method called MR_NET, utilizing convolutional neural networks, to detect and localize breast cancer from histological images. The study aims to improve early and accurate detection of breast cancer, which is crucial for effective treatment and improved survival rates. The research also introduces explainable predictions through heatmaps, highlighting regions within tissue images that the model focuses on when predicting a label, aiding clinicians in making informed decisions. The study, published in Sensors, evaluates the performance of MR_NET and other deep learning models in detecting benign, atypical, and malignant tumors. [Extracted from the article]