ISSN / eISSN: 0033-8362 / 1826-6983
editor@radiologiamedicajournal.com
ISSN / eISSN: 0033-8362 / 1826-6983
ISSN / eISSN: 0033-8362 / 1826-6983
editor@radiologiamedicajournal.com
ISSN / eISSN: 0033-8362 / 1826-6983
Dr. Julia Steiner¹, Dr. Sophia Mehta², Dr. David Okafor³
1 – Department of Breast Imaging, University Hospital Munich, Germany
2 – Department of Radiology, Tata Memorial Hospital, Mumbai, India
3 – Department of Diagnostic Imaging, University of Lagos, Nigeria
Purpose: To assess the effect of AI-assisted mammography interpretation on radiologist diagnostic accuracy and reading time.
Methods: 10 radiologists reviewed 400 digital mammograms (200 with AI support, 200 without). Accuracy, sensitivity, specificity, and mean interpretation time were compared.
Results: AI-assisted reading improved sensitivity from 86% to 94% and reduced reading time by 27%. Specificity remained stable (92%). The most significant gains were seen in detecting architectural distortions and subtle calcifications.
Conclusion: Collaborative AI integration into mammography reading workflows enhances detection performance and efficiency without compromising specificity.
Keywords: Mammography, artificial intelligence, breast cancer, diagnostic accuracy, workflow efficiency
Please fill in the details below to request access to this article or subscribe for updates. Our team will contact you shortly.