AI in Pathology: Why There Are So Few Cleared Devices
Radiology has hundreds of authorized AI devices. Pathology has a handful. The gap is not about difficulty ... it is about whether the images exist at all.
Medical Specialty Hub
AI in Oncology
Cancer detection, treatment planning support, and drug discovery applications of AI in oncology.
Overview
Oncology is one of the highest-stakes areas for clinical AI. AI tools are supporting tumor detection, pathology analysis, treatment planning, clinical trial matching, and genomic analysis to support more precise cancer care.
Oncology has no AI device category of its own, and that absence is more informative than a list would be.
Cancer care is distributed across imaging, pathology, genomics, and treatment planning, and the authorized devices that serve it are filed under those specialties instead. Lung nodule risk stratification is a radiology device. Prostate biopsy screening is a pathology device. Mammography detection is a radiology device. Oncology is the clinical destination, not the regulatory category.
Where the real work is happening
The genuinely oncology-specific AI ... treatment response prediction, biomarker inference from morphology, therapy selection from multi-omic data ... sits almost entirely in research rather than in authorized devices. That is partly a reference-standard problem: predicting who will respond to a therapy requires outcome data collected over years, and no retrospective image set contains it.
Radiomics in particular has produced a large literature and few reproducible findings, largely because feature counts in the hundreds against small patient cohorts is a reliable recipe for spurious association.
Clinical Use Cases
- Tumor detection in imaging
- Histopathology slide analysis
- Genomic biomarker identification
- Clinical trial matching
- Treatment response prediction
Where The Evidence Actually Stands
The gap between published oncology AI and authorized oncology AI is the widest of any specialty on this site. Read that literature with external validation as the first filter, and treat any prognostic model without prospective outcome data as hypothesis-generating. The devices that do serve cancer care today are detection and risk-stratification tools filed under imaging and pathology.
Related Research
Four Clinical AI Diagnostic Tasks and How To Evaluate Them
Triage, detection, characterization, and autonomous diagnosis make different clinical claims. Here is the evidence question that belongs to each one.
AI in Radiology: Practical Clinical Uses
Radiology has the highest concentration of FDA-cleared AI medical devices. Here is how AI is being used in practice in imaging departments today.