FDA Clears AI-Powered Digital Cytology Platform for Cervical Cancer
Combining advanced volumetric imaging technology with deep learning, the Genius Digital Diagnostics System has reportedly shown a 28 percent reduction in false negatives for high-grade squamous intraepithelial lesions and other severe lesions in comparison to microscopic review.
The Food and Drug Administration (FDA) has granted 510(k) clearance to the Genius Digital Diagnostics System, an artificial intelligence (AI)-enabled digital cytology platform that may facilitate enhanced sensitivity for early diagnosis of cervical cancer.
While traditional cervical cancer screenings have involved microscopic review of samples from Pap smear testing, the Genius Digital Diagnostics System facilitates digital image review of these samples with a Genius Cervical AI algorithm that alerts clinicians to suspicious cells for further review, according to Hologic, the developer of the Genius Digital Diagnostics System.
Combining deep learning capability with advanced volumetric imaging technology, the Genius Digital Diagnostics System may facilitate earlier diagnosis of cervical cancer. (Image courtesy of Hologic.)
In comparison to microscopic review of Pap smear testing, Hologic noted that research has demonstrated a 28 percent reduction in false negatives for high-grade squamous intraepithelial lesions and other severe lesions with the Genius Digital Diagnostics System.
“Our technologies have had a tremendous impact on decreasing cancer rates in women, and we are incredibly excited by the promise of Genius Digital Diagnostics. The system delivers more actionable and accurate insights for laboratories and healthcare professionals to enhance patient care,” said Jennifer Schneiders, Ph.D., the president of diagnostic solutions at Hologic.
Considering Breast- and Lesion-Level Assessments with Mammography AI: What New Research Reveals
June 27th 2025While there was a decline of AUC for mammography AI software from breast-level assessments to lesion-level evaluation, the authors of a new study, involving 1,200 women, found that AI offered over a seven percent higher AUC for lesion-level interpretation in comparison to unassisted expert readers.
