What happened
MediSpectra of Lexington, Massachusetts, developed the LUMA Cervical Imaging System. It scans the cervix with fluorescence and reflected light during colposcopy. A logistic classifier and a multivariate analysis, optimized to separate high-grade lesions (CIN 2,3+) from other tissue, color a map of the cervix. In a study at seven centers, 16 colposcopists committed to their own biopsies before they viewed the map. The study stopped early at 227 of a planned 788 women. Of 193 evaluable women, colposcopy found CIN 2,3+ in 41, and LUMA-directed biopsies found 9 more. FDA reviewers report that this was a 22.0% relative gain. LUMA alone sent 18.1% of women to biopsies that found no high-grade disease, above the 15% goal. A larger randomized study found no statistically significant gain, and an advisory panel recommended against approval. FDA approved the system on 16 March 2006 and judged that a trade-off of one extra case for about four unnecessary biopsies supported approval.
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), P040028: LUMA Cervical Imaging System. Approved 2006-03-16.
where the LUMA System was used in sequence after colposcopy, showed a relative gain of approximately 22.0% (95% CI 6.1-37.8) in the identification of true-positive cases of high-grade CIN 2,3+ across the study population.
Details
AI system
- Workflow
-
Colposcopists committed to their own biopsy sites before they viewed the LUMA map. The probe scanned the cervix for about 12 seconds after acetic acid was applied.
Follow-up
- Conditions for replication
-
The woman was referred to colposcopy after an abnormal Pap test. LUMA is used as an adjunct to colposcopy.
Entry details
- Sector
-
Healthcare.
- Purpose
-
MediSpectra built the LUMA Cervical Imaging System to help colposcopists identify high-grade cervical disease in women referred after an abnormal Pap test.
Not yet checked: model, release, provider, role, configuration, inputs and outputs, autonomy, human oversight, AI contribution, 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). LUMA imaging found more high-grade cervical lesions after colposcopy (AIBD-0796). Artificial Intelligence Benefits Database, edition of 31 October 2026. https://petervartanian.xyz/aibd/records/AIBD-0796.html (archived at https://doi.org/10.7910/DVN/AADSII, version 4.0)