{"entity":{"id":"paper-wu-nat-med","kind":"paper","name":"How medical AI devices are evaluated: limitations and recommendations from an analysis of FDA approvals","aka":[],"tldr":"Paper cited by one bottleneck page and 22 idea pages, indexed on Europe PMC as PubMed record 33820998 and published in Nature Medicine; the citing pages link this DOI, which is how the record was matched.","summary":"Europe PMC indexes no abstract for this record; the title is the only text available.\n\nIndexed on Europe PMC as PubMed record 33820998 (DOI 10.1038/s41591-021-01312-x). Matched by DOI alone: one bottleneck page and 22 idea pages cite this DOI among their external links (the pages are listed under Related), and this page was written so that the citation resolves inside OnCo. No figure has been checked by an editor.","asOf":"2026-09-22","links":[{"label":"Nat Med 2021","url":"https://doi.org/10.1038/s41591-021-01312-x"},{"label":"PubMed","url":"https://pubmed.ncbi.nlm.nih.gov/33820998/"},{"label":"Europe PMC","url":"https://europepmc.org/article/MED/33820998"}],"tags":["europepmc-ingest"],"related":["b-ai-validation","idea-data-ai-liability-safe-harbour","idea-data-silent-trial-before-deployment","idea-data-neutral-ai-evaluator","idea-data-precompetitive-cancer-foundation-model","idea-data-llm-documentation-rct-oncology","idea-data-ai-vs-tumour-board-rct","idea-data-ai-pathology-evaluation-standard","idea-data-drift-monitoring-standard","idea-moon-continuous-ai-validation-registry","idea-data-digital-twin-predict-then-observe","idea-data-ai-audit-trail-in-ehr","idea-data-multisite-validation-precondition","idea-data-in-silico-trials-calibrated","idea-data-ai-post-market-performance-reporting","idea-data-subgroup-performance-reporting-mandate","idea-data-patient-facing-model-cards","idea-data-ai-reimbursement-tied-to-outcomes","idea-data-open-weights-for-public-funded-ai","idea-data-ai-red-team-programme","idea-data-ai-screening-endpoints","idea-data-ai-decommissioning-rules","idea-moon-validated-digital-twins"],"cancers":[],"sections":[],"technologies":[],"targets":[],"drugs":[],"companies":[],"institutions":[],"pathways":[],"terms":[],"trials":[],"people":[],"bottlenecks":[],"keyPapers":[],"journals":["nature-medicine"],"dependsOn":[],"notes":[],"journal":"Nature Medicine","year":2021,"doi":"10.1038/s41591-021-01312-x","pmid":"33820998","authors":"Wu E, Wu K, Daneshjou R, et al.","paperType":"observational","findings":[],"whatItMeans":"One bottleneck page and 22 idea pages on OnCo cite this paper by its DOI; this record gives the citation a page of its own so a reader can follow it without leaving OnCo. Read the abstract above alongside the citing pages listed under Related; the record was created automatically from the Europe PMC entry and its figures have not been checked by hand.","caveats":["Matched to the citing OnCo records by DOI alone; the summary reproduces the Europe PMC abstract and no figure has been verified against the full paper."]},"route":"/key-papers/paper-wu-nat-med/","neighbours":{"bottleneck":[{"id":"b-ai-validation","kind":"bottleneck","name":"AI that is built but not validated or deployed","route":"/bottlenecks/b-ai-validation/"}],"idea":[{"id":"idea-data-ai-liability-safe-harbour","kind":"idea","name":"A liability framework for clinical AI: safe harbour for clinicians, liability for makers","route":"/ideas/idea-data-ai-liability-safe-harbour/"},{"id":"idea-data-silent-trial-before-deployment","kind":"idea","name":"A mandatory silent (shadow) trial before any cancer AI goes live","route":"/ideas/idea-data-silent-trial-before-deployment/"},{"id":"idea-data-neutral-ai-evaluator","kind":"idea","name":"A neutral public evaluator for cancer AI, on the model of NIST","route":"/ideas/idea-data-neutral-ai-evaluator/"},{"id":"idea-data-precompetitive-cancer-foundation-model","kind":"idea","name":"A pre-competitive consortium to train a shared multimodal cancer foundation model","route":"/ideas/idea-data-precompetitive-cancer-foundation-model/"},{"id":"idea-data-llm-documentation-rct-oncology","kind":"idea","name":"A randomised trial of AI scribes in oncology clinics measuring errors and time","route":"/ideas/idea-data-llm-documentation-rct-oncology/"},{"id":"idea-data-ai-vs-tumour-board-rct","kind":"idea","name":"A randomised trial of AI-generated treatment recommendations versus tumour boards","route":"/ideas/idea-data-ai-vs-tumour-board-rct/"},{"id":"idea-data-ai-pathology-evaluation-standard","kind":"idea","name":"A standard evaluation pathway for AI-assisted pathology, from reader study to deployment","route":"/ideas/idea-data-ai-pathology-evaluation-standard/"},{"id":"idea-data-drift-monitoring-standard","kind":"idea","name":"A standard for monitoring AI performance drift with pause thresholds","route":"/ideas/idea-data-drift-monitoring-standard/"},{"id":"idea-moon-continuous-ai-validation-registry","kind":"idea","name":"Continuous prospective validation for every oncology AI tool after deployment","route":"/ideas/idea-moon-continuous-ai-validation-registry/"},{"id":"idea-data-digital-twin-predict-then-observe","kind":"idea","name":"Digital twins for treatment selection, validated by predicting before observing","route":"/ideas/idea-data-digital-twin-predict-then-observe/"},{"id":"idea-data-ai-audit-trail-in-ehr","kind":"idea","name":"Every AI output logged in the record with input hash, version and clinician response","route":"/ideas/idea-data-ai-audit-trail-in-ehr/"},{"id":"idea-data-multisite-validation-precondition","kind":"idea","name":"External validation at five or more sites in two countries before clearance","route":"/ideas/idea-data-multisite-validation-precondition/"},{"id":"idea-data-in-silico-trials-calibrated","kind":"idea","name":"In silico trials to prioritise combinations, scored against later real trials","route":"/ideas/idea-data-in-silico-trials-calibrated/"},{"id":"idea-data-ai-post-market-performance-reporting","kind":"idea","name":"Mandatory post-market performance reporting for cancer AI","route":"/ideas/idea-data-ai-post-market-performance-reporting/"},{"id":"idea-data-subgroup-performance-reporting-mandate","kind":"idea","name":"Mandatory subgroup performance reporting for cancer AI","route":"/ideas/idea-data-subgroup-performance-reporting-mandate/"},{"id":"idea-data-patient-facing-model-cards","kind":"idea","name":"Patients told which AI is used in their care, in plain language","route":"/ideas/idea-data-patient-facing-model-cards/"},{"id":"idea-data-ai-reimbursement-tied-to-outcomes","kind":"idea","name":"Pay for cancer AI only when it has outcome evidence, then pay properly","route":"/ideas/idea-data-ai-reimbursement-tied-to-outcomes/"},{"id":"idea-data-open-weights-for-public-funded-ai","kind":"idea","name":"Publicly funded cancer AI must release open weights and model cards","route":"/ideas/idea-data-open-weights-for-public-funded-ai/"},{"id":"idea-data-ai-red-team-programme","kind":"idea","name":"Red-team programmes that attack cancer AI before patients do","route":"/ideas/idea-data-ai-red-team-programme/"},{"id":"idea-data-ai-screening-endpoints","kind":"idea","name":"Require stage-shift or interval-cancer endpoints for AI in cancer screening","route":"/ideas/idea-data-ai-screening-endpoints/"},{"id":"idea-data-ai-decommissioning-rules","kind":"idea","name":"Rules for retiring cancer AI when performance drops or the standard of care moves","route":"/ideas/idea-data-ai-decommissioning-rules/"},{"id":"idea-moon-validated-digital-twins","kind":"idea","name":"Whole-patient digital twins validated in prospective randomised trials","route":"/ideas/idea-moon-validated-digital-twins/"}],"journal":[{"id":"nature-medicine","kind":"journal","name":"Nature Medicine","route":"/journals/nature-medicine/"}]}}