AI in Cancer Detection and Diagnosis

The implementation of advanced artificial intelligence (AI) algorithms, in particular deep learning, into clinical practice has the potential to propel forward the field of oncological radiology and ultimately improve patient treatment.

Upvote 0 Comment

We are currently witnessing a big advance in the ability of artificial intelligence (AI) algorithms, in particular deep learning, to interpret sensory information, enabling machines to better recognise and represent complex patterns in imaging data. This progress means that certain AI algorithms are now able to match or in some cases, even outperform humans in narrow tasks. Within radiology, trained physicians visually evaluate medical images in order to detect, diagnose and monitor cancer in patients. Such evaluation is frequently based on experience and can, on occasion, be subjective. Hugo J.W.L. Aerts and colleagues suggest that integrating advanced AI into clinical practice could be a paradigm shift supporting radiologists in cancer detection and diagnosis by providing quantitative and automated assessments of medical images and ultimately improving patient treatment. In this Opinion article, “Artificial Intelligence in Radiology,” in Nature Reviews Cancer, the authors discuss the application of AI to image-based tasks in the field of oncological radiology and consider the advantages and challenges of its clinical implementation.


By Anna Dart, Nature Reviews Cancer

Source: https://rdcu.be/OJca

Image credits: Hosny et al. Nature Reviews Cancer (2018) doi:10.1038/s41568-018-0016-5

No comments yet.