Pathology · Prostate biopsy AI detection
Paige Prostate Detect
Paige.AI (acquired by Tempus AI, August 2025)
The first FDA-authorised AI in digital pathology (De Novo, September 2021), Paige Prostate Detect anchors the regulatory pioneer status of the field — pivotal clinical validation across 218 institutions and 610 whole slide images demonstrated a 7.9-percentage-point sensitivity gain (88.7% → 96.6%) when pathologists worked with AI assistance, with a 70% reduction in false-negative diagnoses.
Performance Metrics
Clinical Evidence
The pivotal study (Raciti et al., *Archives of Pathology & Laboratory Medicine*, October 2023; n=610 WSIs from 218 institutions; 18 pathologists including 2 genitourinary subspecialists) established the clinical benefit case for FDA De Novo authorisation. Pathologists reviewed each WSI twice — first unassisted, then with Paige Prostate Detect — in a prospective reader-study design. Average pathologist sensitivity rose from 88.7% to 96.6% (specificity: 97.3% → 98.0%), a 7.9 pp absolute gain. *PaPr correctly classified 100% of WSIs on which pathologists changed their diagnosis.* The result implies a 70% reduction in false-negative diagnoses and a 24% reduction in false-positive diagnoses compared to unassisted reads. An independent real-world validation by da Silva et al. (*The Journal of Pathology*, 2021; n=100 consecutive patients, 682 part-specimens, external laboratory in a country not involved in Paige's development) reported standalone sensitivity of 0.99 (95% CI 0.96–1.0), specificity 0.93 (95% CI 0.90–0.96), and NPV 1.0 (95% CI 0.98–1.0) at the part-specimen level; patient-level sensitivity was 1.0 (95% CI 0.93–1.0). Diagnostic time was estimated to fall by 65.5% with AI assistance. At the patient level, three malignant cases were missed unassisted but detected with AI. The CONFIDENT P trial (Hendriks et al., *JCO Clinical Cancer Informatics*, March 2025; NCT-registered prospective implementation study; n=82 patients / 237 slides; University Medical Center Utrecht) evaluated AI-guided selective immunohistochemistry (IHC). AI assistance reduced the relative risk of IHC use per detected cancer case to 0.55 at the patient level (RR 0.41 at slide level), with €1,700 IHC cost savings in the trial cohort. Pathologist diagnostic confidence was significantly higher in the AI arm (80% vs 56% confident or high-confidence reads). A NICE Medtech Innovation Briefing (MIB280) assessed the evidence base and reported strong standalone performance corroborated by the da Silva independent validation. NICE noted that further prospective real-world studies are warranted before widespread NHS adoption. Following Tempus AI's acquisition of Paige in August 2025, Paige Prostate Detect is now deployed under the Tempus digital pathology portfolio; the approved indications and algorithm remain unchanged.
| Study | Design | n | Sensitivity | Specificity | AUC | Published |
|---|---|---|---|---|---|---|
| Raciti P, Sue J, Ceballos R, et al. (Paige.AI / Mount Sinai) | ProspectiveProspective | 610 | 96.6% | 98.0% | — | Archives of Pathology & Laboratory Medicine, Oct 2023 |
| da Silva LM, Pereira EM, Salles PGO, et al. | RetrospectiveRetrospective | 100 | 99% | 93% | — | The Journal of Pathology, Jun 2021 |
| Hendriks RJL, van der Leest MMG, Vergunst H, et al. (CONFIDENT P) | ProspectiveProspective | 82 | — | — | — | JCO Clinical Cancer Informatics, Mar 2025 |
Clinical Pulse
Verified clinician sentiment
Verified clinician reviews launching · Apply to contribute →
Reviews are signed by clinicians verified against their regulator's public register.
Inside the Auris+ Listing
Six more sections complete this device’s Auris+ Listing.
Inside the algorithm
ProPro unlocks the five-stage editorial walkthrough of how this device's AI reaches a verdict — model architecture, training data, inference latency, and where the clinician stays in the loop.
Decision Ledger
ProPro unlocks a private, cross-vendor log of every case you read on this device — what the AI called, what you concluded, and a one-line reason if you overrode it. Ready for the EU AI Act's deployer logging obligations when they land in 2028.
Clinical Evidence Deep Dive
ProPro unlocks the structured clinical-evidence summary — study count, target patient population, and a tabular accuracy-metrics view drawn from peer-reviewed sources.
Peer-Reviewed Publications
ProPro unlocks the curated peer-reviewed publication list with PubMed cross-links — the citation backbone of every editorial verdict.
Post-Market & Regulatory Conditions
ProPro unlocks the post-market surveillance summary, recall record, and the conditions of approval that bound real-world use.
AI Algorithm Version History
ProPro unlocks the chronological record of algorithm version changes — what changed when, drawn from manufacturer changelogs and regulatory filings.
Regulatory Approvals
Class C
DEN200080
Class II
Safety Record
No device recalls, Class I/II FDA safety communications, or significant MAUDE adverse-event clusters specific to Paige Prostate Detect have been identified in publicly available databases as of the review date. The device is a software-only second-read overlay that does not alter slide preparation, staining, or tissue handling, which substantially limits the harm surface. The system's single-coordinate output is intended to direct — not replace — pathologist review; final diagnosis remains with the treating pathologist. The CONFIDENT P trial reported no diagnostic safety incidents attributable to AI-guided selective IHC. Procurement officers should note that the device now operates under Tempus AI's quality management system following the August 2025 acquisition; continuity of regulatory status has been confirmed (EU IVDR certification transferred in 2025).
Intended Use & Indications
Paige Prostate Detect is a software-only device indicated to assist qualified pathologists in reviewing scanned whole slide images (WSI) from prostate needle core biopsies. After the pathologist completes their initial diagnostic review of the WSI, if the algorithm detects tissue morphology suspicious for malignancy, it surfaces a single X/Y coordinate indicating the location with the highest estimated cancer probability. The system delivers a WSI-level binary output (suspicious / benign) together with a spatial pointer to guide re-review. All diagnostic and clinical decisions remain with the treating pathologist.