What happened
QView Medical of Los Altos, California, makes QVCAD, software that marks suspicious areas in 3D automated breast ultrasound. Its engine uses pattern recognition and artificial neural networks trained on biopsy-proven cancers, benign lesions and normal cases. In a retrospective reader study, 18 radiologists from practices in nine US states read 185 cases with and without the software. The cases came from women with dense breasts screened at 13 US sites in 2009 and 2010, and 52 had cancer. FDA reviewers report that reading time fell from 215 to 144 seconds per case, a 71-second or 33% reduction. The average AUC was 0.7353 with the software and 0.7493 without it, which met the non-inferiority margin. FDA approved QVCAD on 9 November 2016. The reviewers write that faster reading may raise patient throughput and help wider use of breast ultrasound. They also warn that readers may rely too much on the absence of marks.
Sources
-
Regulatory document, by authors with no tie to the system.
U.S. Food and Drug Administration, Center for Devices and Radiological Health. Summary of Safety and Effectiveness Data (SSED), P150043: QVCAD System. Approved 2016-11-09.
These results show that the estimated ∆Time per case is 71 seconds (33% of the average TimeABUS; No QVCAD), which is statistically significantly greater than zero (p=0.000016).
Details
AI system
- Workflow
-
Radiologists completed eight hours of hands-on training with the software. Readers reviewed the original ultrasound images alongside the marked projection image.
Follow-up
- Conditions for replication
-
The images come from the somo•v Automated Breast Ultrasound System. The woman has dense breasts and a screening mammogram rated BI-RADS 1 or 2.
Entry details
- Sector
-
Healthcare.
- Purpose
-
QView Medical built QVCAD to help radiologists search 3D ultrasound images of dense breasts for lesions that mammograms do not show.
Not yet checked: model, release, provider, role, configuration, inputs and outputs, autonomy, human oversight, beneficiaries, reach, distribution and access, duration, every measurement question, every money question, ownership and control, dependencies, cross-border connections, geopolitical implications, replication attempts, availability, institutional decision, grounds, action and result, success criterion, every evidence question, observed or intended, alternative assessments, related entries, other identifiers, category mappings.
Cite this record
Peter H. Vartanian. (2026). QVCAD cut reading time for breast ultrasound by a third (AIBD-0793). Artificial Intelligence Benefits Database, edition of 31 October 2026. https://petervartanian.xyz/aibd/records/AIBD-0793.html (archived at https://doi.org/10.7910/DVN/AADSII, version 4.0)