Search

A study conducted by researchers at the Cyprus Marine and Maritime Institute explores the use of machine learning in digital pathology for diagnosis, prognosis, and personalization. The researchers introduce a new approach called magnifying networks (MagNets) to address the challenge of analyzing extremely large digitized images. By using an attention mechanism, MagNets identify regions of the image that require a finer analysis, resulting in a more efficient classification process. The study demonstrates the effectiveness of MagNets in whole-slide image classification and highlights their ability to process fewer patches compared to existing approaches. [Extracted from the article]