AI biopsy review catches more prostate cancers in HTW's pooled studies

Health Technology Wales found that pathologists' sensitivity for prostate cancer rose from 91.6% to 96% with AI help in four pooled studies. It judged the evidence insufficient for routine use.

  • 2 for an organization

    In HTW's pooled analysis, pathologists' sensitivity for cancer rose from 91.6% to 96% with AI help, while specificity did not change.

Depth runs from 1, a convenience or time saved, to 5, a life or a livelihood preserved.

Status
Partly reported benefit.
Evidence

foundstudiedreportedclaimed

Two independent assessment reports support the benefit.

Compared with

Pathologists reviewing the same biopsy slides without AI, pooled across four studies. Measured.

Without the system
91.6%
With it
96%
The system’s part
Every study that HTW found compared pathologists reviewing slides alone with pathologists aided by AI, which isolates the software's part. HTW describes all three tools as convolutional neural networks trained on whole slide images. Most studies were retrospective, and several were not peer reviewed.
Not yet checked
LimitsCosts and harmsLearns from dataLater outcome

Download data Cite Suggest a correction

Benefit type
Improved Task Accuracy
System
AI-assisted review of prostate biopsies: Galen Prostate, Paige Prostate and DeepDx Prostate
Made by
Ibex Medical Analytics (Galen Prostate), Paige AI (Paige Prostate) and Deep Bio (DeepDx Prostate)
Used by
Pathology services in the studies that HTW reviewed, with Galen Prostate introduced across NHS Wales through a Welsh Government innovation fund pilot
Where
Multiple countries (studies), with guidance for Wales, United Kingdom
Stage
In a pilot
Date
, when the source was published
Kind
Aggregate result
Broad benefit
Identifying Objects, Actions, and Events, Judging the Qualities of Objects, Services, or People
Collections
Scrutiny and findings
Last revised
1 November 2026

Sources

  1. Assessment report, by authors with no tie to the system.

    Health Technology Wales. HTW Guidance 057: Artificial Intelligence (AI) assisted review of prostate biopsies in the detection and diagnosis of prostate cancer. May 2024, Evidence summary.

    Overall, AI was found to numerically improve the sensitivity of diagnosis, without impacting negatively on specificity, indicating there was a reduction in false negatives with no increase in false positives.

  2. Assessment report, by authors with no tie to the system.

    Health Technology Wales. Evidence Appraisal Report EAR057: Artificial Intelligence (AI) assisted review of prostate biopsies in the detection and diagnosis of prostate cancer. May 2024, section 5.2 and Table 2.

    We carried out our own pooled analysis of sensitivity which indicated that AI-assistance improved sensitivity for cancer detection by around 4.4% from 91.6% to 96%, p=0.001 (Table 2).

Details

AI system

Workflow

Pathologists reviewed digitized whole slide images while the AI flagged suspicious areas and suggested Gleason grades. The studies compared each pathologist's review with and without AI help on the same slides.

Follow-up

Conditions for replication

Slides are scanned to good quality, which experts told HTW matters most. The AI is integrated with local laboratory information systems and has the support of several local teams.

Entry details

Sector

Healthcare.

Purpose

Ibex Medical Analytics, Paige AI and Deep Bio built Galen Prostate, Paige Prostate and DeepDx Prostate to flag cancer on digitized prostate biopsy slides and help pathologists grade it.

30 more questions 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). AI biopsy review catches more prostate cancers in HTW's pooled studies (AIBD-1121). Artificial Intelligence Benefits Database, edition of 1 November 2026. https://petervartanian.xyz/aibd/records/AIBD-1121.html (archived at https://doi.org/10.7910/DVN/AADSII, version 7.0)

BibTeX Data Field definitions