Observed arrival · 2026-10-03
Orthoharness asks whether clinical AI can show its work
An open-source reference harness for evaluating orthopaedic AI with staged cases, separate clinical yardsticks, and checks for omissions, citations, and decision process.
- For
- Orthopaedic clinicians and teams evaluating clinical AI
- Worth noticing
- Seven demo cases are synthetic, and the page labels the implementation unvalidated; its loss weights are placeholders.
Field notes
The page organizes its case for a specialty-specific harness around nine gaps in existing tests, then pairs each gap with a proposed scoring or record-keeping check. It distinguishes citation existence from whether a source supports a claim, reserving the latter for human review. The implementation is explicitly a reference version: demo cases are synthetic, the loss weights are placeholders, and the site says neither is clinical guidance.
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
One card from the complete issue