How Do Perceptions of Radiologists and AI Systems in Breast Screening Differ?

How Do Perceptions of Radiologists and AI Systems in Breast Screening Differ?
share Share

A study published in Nature Scientific Reports revealed striking differences in the way radiologists and artificial intelligence (AI) systems perceive breast cancer lesions. Given that how they recognise breast cancer differs, this New York University study demonstrates the potential value of combining both human and AI methods when making medical diagnoses.

 

The AI system in question is a deep neural network (DNN) which simulates the learning performed by neurons. On a data set consisting of 229,426 screening mammography exams from 141,473 patients, this study’s DNNs learned to distinguish cancerous lesions since each exam was marked as malignant or benign. One issue with the approach is that it is sometimes different to understand what features the AI uses to differentiate. Thus, in the current study, seven radiologists and an AI read the same 120 exams. Both the diagnosis and features used by radiologists and AI were compared.

 



Radiologists diverged significantly from the AIs in how they diagnosed soft tissue lesions. While radiologists relied on brightness and shape, the DNNs depended on high-frequency components scattered across the images and outside regions radiologists considered suspicious.

 

“In these breast-cancer screenings, AI systems consider tiny details in mammograms that are seen as irrelevant by radiologists,” explains Dr Krzysztof J. Geras, the study’s senior author. ‘This divergence in readings must be understood and corrected before we can trust AI systems to help make life-critical medical decisions.’


Click here for the latest Women's Health news 

 

Source: Nature Scientific Reports, Eurekalert!

Image Credit: iStock

 

«« How Good is the Evidence Informing Breast Cancer Clinical Practice Guidelines?


ASCO’s Clinical Practice Guideline Now Supports Extended Endocrine Therapy with Breast Cancer Index™ »»

Published on : Thu, 5 May 2022



Related Articles

A study published in Clinical Breast Cancer  revealed that the quality of the systematic reviews that provide the evidence to... Read more

Flatiron Data Supports Using Palbociclib Plus an Aromatase Inhibitor for Metastatic Breast Cancer

This Pfizer-sponsored study used de-identified data derived from the Flatiron Health Analytic Database in a retrospective observational... Read more

AI Catches What Radiologists Miss on Breast Screening

A large German study found that AI can accurately detect and localise breast cancers on screening exams missed by radiologists,... Read more

breast imaging, breast screening, Radiologists A study published in Nature Scientific Reports revealed striking differences in the way radiologists and artificial intelligence (AI) systems perceive breast cancer lesions.

No comment


Please login to leave a comment...

Highlighted Products