# AI-first reading for high-volume common cancer diagnoses, pathologist for the exceptions

Source: https://onco.cc/ideas/idea-acc-ai-first-pathology-common-cases/  
OnCo record `idea-acc-ai-first-pathology-common-cases` (Idea). Data CC BY-NC 4.0, attribute "Data from OnCo (onco.cc)"; commercial use needs a licence.

## TL;DR

Let validated AI make the first read on routine, high-volume samples like cervical smears and standard breast biopsy stains, so scarce pathologists spend their time on the difficult cases.

## Summary

Pathology AI is now good enough for narrow, high-volume tasks: cervical cytology screening, prostate biopsy detection, HER2 and ER scoring on breast biopsies. In systems with a fraction of the needed pathologists, an AI-first workflow with pathologist sign-off only on flagged or discordant cases could multiply capacity. The unsolved problems are local validation on different scanners and populations, regulatory acceptance in each jurisdiction, and liability.

## Fields

- Kind: Idea
- Last checked: 2026-09-08
- Hypothesis: In a pathologist-scarce setting, an AI-first workflow will at least triple cases reported per pathologist-hour while keeping sensitivity for malignancy above 98% on prospective audit.
- Rationale: AI cervical screening tools have already shown non-inferiority to cytotechnologists in several settings; the same triage logic is used for tuberculosis chest X-rays in high-burden countries.
- Proposed test: A prospective, paired-read study in two LMIC laboratories on cervical cytology and breast core biopsies: AI-first with sign-off versus full human read, measuring throughput, sensitivity, specificity, and time to report.
- Maturity: early-clinical
- Actor: data

## Sources

- Bottleneck evidence (Most of the world has almost no cancer care): WHO fact sheet, Cancer: https://www.who.int/news-room/fact-sheets/detail/cancer

## Connected records

- ideas: [AI quantification of HER2-low and HER2-ultralow](https://onco.cc/ideas/idea-ai-her2-low-scoring/), [Pathologist assistants plus AI triage to multiply pathologist capacity](https://onco.cc/ideas/idea-acc-pathologist-assistants-and-ai-triage/)
- cancers: [Cervical cancer](https://onco.cc/cancers/cervical/), [HR-positive / HER2-negative breast cancer](https://onco.cc/cancers/breast-hr-positive/)
- technologies: [Digital pathology & AI](https://onco.cc/technologies/digital-pathology-ai/), [Pathology & radiology foundation models](https://onco.cc/technologies/pathology-foundation-model/)
- companies: [Lunit](https://onco.cc/companies/lunit/), [Owkin](https://onco.cc/companies/owkin/), [Paige AI](https://onco.cc/companies/paige/), [PathAI](https://onco.cc/companies/pathai/)
- bottlenecks: [AI that is built but not validated or deployed](https://onco.cc/bottlenecks/b-ai-validation/), [Most of the world has almost no cancer care](https://onco.cc/bottlenecks/b-global-access/), [Not enough oncologists, nurses, pathologists, physicists](https://onco.cc/bottlenecks/b-workforce/)
- roadmaps: [AI in oncology roadmap: pattern readers → foundation models → agents in the workflow](https://onco.cc/roadmaps/ai-oncology-roadmap/)

---
JSON: https://onco.cc/api/v1/entities/idea-acc-ai-first-pathology-common-cases.json