Lunit INSIGHT DBT (V1.2)

Evidence summary

FDA K253796 packet reports 3,277 DBT exams from US healthcare institutions. Standalone results: ROC AUC 0.9388 (95% CI 0.9304-0.9472), JAFROC AUC 0.9206 (0.9117-0.9295), and operating-point sensitivity/specificity of 91.11%/77.62% at 0.1, 88.38%/83.68% at 0.3, and 81.48%/93.44% at 0.6. Three-way lesion-type agreement was 75.61% (73.40%-77.80%). A 258-case reader comparison reported AI standalone AUROC 0.9430 versus average reader 0.8983.

• FDA source
• Exact FDA submission
• Checked 2026-09-08

• Source

No contextualized exact-version metric has completed evidence review. Regulatory-document values, when available, are shown above with their limitations.

Postmarket safety review

The exact-identifier FDA recall query returned no native recall record for K253796. This limited result does not establish absence of recalls or other safety information and does not cover later versions, family records, MAUDE reports, corrections, or field notices.

• Public source
• Exact FDA submission
• Checked 2026-09-08

• Source

• Verify DICOM, HL7/FHIR, PACS/RIS/EHR integration: Attach protocol-level DICOM, HL7, FHIR, PACS, RIS, EHR, or orchestrator documentation.
• Verify deployment and data flow: The reviewed sources do not establish the current exact-release hosting topology, data flow, region, tenancy, or disaster-recovery configuration.
• Verify security and privacy controls: Exact-release security controls, data retention, hosting boundaries, and scoped assurance documentation require vendor confirmation.
• Verify training: Current training prerequisites, competency checks, and support commitments for the exact configuration require vendor confirmation.
• Verify pricing and total cost: No public price or quote for the exact K253796 configuration was established in this review.
• Verify reimbursement and coding relevance: No exact-product payer policy, coding instruction, or payment determination was established; eligibility must be evaluated by payer, site of service, and use case.
• Verify local validation and lifecycle monitoring: Current drift, quality, uptime, alert, escalation, and incident-response commitments for the exact configuration require vendor confirmation.

Exact-submission ACR model-card fields were normalized under the current provenance rules; vendor and independent-study confirmation remain distinct. Exact FDA scope controls clinical claims. ACR provides bounded exact-submission catalog context. Candidate literature metadata remains a discovery queue until full-text identity, endpoint, population, and version review is completed. Automated exact-name discovery found 0 PubMed and 1 ClinicalTrials.gov candidate records. Candidates require human product and version matching; zero candidates is not evidence that no studies exist. Native FDA recall identifiers produced 0 postmarket candidate records; 0 have been reviewed (0 published, 0 rejected) and 0 remain unreviewed.

Candidate leads remain unpublished until a human confirms the exact product and tested version.

Source classes: fda ai list, acr ai central product, fda decision summary, fda device recall, literature index, trial registry, pubmed, clinical trials, openfda device recall