Breast Case Scenario

Breast cancer is the most common cancer in the UK, accounting for 15% of all new cancer cases. If breast cancer is spotted earlier, then the patient’s chances of survival dramatically increase. Mammograms are a well-known way that medics can detect and measure breast density and spot any cancers that may be starting to form. But this analysis is done manually so it can be misinterpreted, different doctors may have varying opinions, it's time consuming and the poor image quality makes small changes difficult to spot with the naked eye. In order to try and make critical diagnoses earlier, researchers applied a machine learning model to analyse mammography images to improve the detection and classification of benign (left) and malignant breast tumours (right), which the model did with incredible accuracy. It was a promising performance that helps the clinician with speedy diagnosis, treatment planning, and follow-up of disease progression.

Written by Sophie Arthur

Sophie Arthur

former Science Communications Officer MRC LMS

Formerly a stem cell biologist, Sophie has since moved into science communication working at the MRC LMS. She also writes an award winning science and education blog Soph talks science.

Website

• Image from work by Tariq Mahmood and colleagues
• Faculty of Information Technology, Beijing University of Technology, Beijing, China and Division of Science and Technology, Department of Information Sciences, University of Education, Lahore, Pakistan
• Image originally published with a Creative Commons Attribution 4.0 International (CC BY 4.0)
• Published in PLOS ONE, January 2022