Medical technology firm Hologic has released new experimental data strengthening the role of artificial intelligence (AI) in breast cancer screening. Conducted at Massachusetts General Hospital, the study retrospectively analyzed 7,500 digital 3D mammograms using the Genius AI platform.
The findings highlight the technology’s ability to “bridge the gap” in clinical diagnosis. Specifically, among 100 documented false-negative cases—where patients actually had cancer despite initial normal screenings—the AI system flagged and accurately localized suspicious areas in approximately one-third of them. Furthermore, the algorithm successfully identified 90% of the cancer cases previously confirmed by radiologists.
The research indicates that the AI is particularly adept at detecting invasive ductal carcinomas and lymph-node-positive cancers. While some limitations remain regarding lobular or low-grade invasive carcinomas, the integration of this technology is viewed as a significant optimization for radiological workflows. Combining human expertise with machine precision not only improves early detection rates but also streamlines scan-reading times, ultimately reducing the burden on healthcare systems.
Source: https://www.fiercebiotech.com/medtech/latest-hologic-breast-cancer-study-bolsters-use-ai-detection