Oncology Eligibility Criteria Matcher
AI-assisted oncology trial matching that extracts biomarker and TNM staging data from unstructured charts and performs transparent criterion-level inclusion and exclusion eligibility assessment.
The Problem
“AI-assisted oncology trial matching with biomarker and TNM extraction plus transparent criterion-level eligibility assessment”
Organizations face these key challenges:
Biomarker and staging data are scattered across pathology, genomics, radiology, and progress notes
Oncology terminology is highly specialized and context dependent
Trial criteria contain nested logic, temporal conditions, and exclusions
Manual abstraction is slow and varies by reviewer experience
Impact When Solved
The Shift
Human Does
- •Review pathology, genomics, radiology, and clinic notes to find biomarker, TNM stage, prior therapy, and performance status details
- •Interpret protocol inclusion and exclusion criteria and compare patient facts against each rule manually
- •Reconcile conflicting or missing chart evidence and decide what information is sufficient for screening
- •Document eligibility findings in tracking fields or spreadsheets and prepare cases for coordinator follow-up
Automation
- •No AI-driven extraction or criterion assessment is used in the legacy workflow
- •No automated consolidation of oncology evidence across unstructured documents is available
- •No system-generated criterion-level pass, fail, or needs-review outputs with citations are produced
Human Does
- •Confirm or correct extracted biomarker, staging, therapy history, and performance status values when confidence is low or evidence conflicts
- •Review criterion-level eligibility results and make final screening and escalation decisions for each patient-trial match
- •Resolve exceptions such as ambiguous staging language, missing records, or protocol interpretation edge cases
AI Handles
- •Extract oncology-specific structured facts from unstructured charts and link each value to supporting source evidence
- •Ingest trial criteria and evaluate each inclusion and exclusion rule with pass, fail, or needs-review outputs
- •Flag missing, conflicting, or time-sensitive data that could change eligibility status
- •Prioritize likely trial matches and generate evidence-backed summaries for coordinator review
Operating Intelligence
How it works
AI runs the first three steps autonomously.
Humans own every decision.
The system gets smarter each cycle.
Who is in control at each step
Each column marks the operating owner for that step. AI-led actions sit above the divider, human decisions and feedback loops sit below it.
Step 1
Assemble Context
Step 2
Analyze
Step 3
Recommend
Step 4
Human Decision
Step 5
Execute
Step 6
Feedback
AI lead
Autonomous execution
Human lead
Approval, override, feedback
AI handles assembly, analysis, and execution. The human gate sits at the decision point. Every cycle refines future recommendations.
The Loop
6 steps
Assemble Context
Combine the relevant records, signals, and constraints.
Analyze
Evaluate options, risk, and likely outcomes.
Recommend
Present a ranked recommendation with supporting rationale.
Human Decision
A human accepts, edits, or rejects the recommendation.
Authority gates · 1
The application must not make the final patient-trial eligibility decision without coordinator review and approval. [S1][S2]
Why this step is human
The decision carries real-world consequences that require professional judgment and accountability.
Execute
Carry out the approved action in the operating workflow.
Feedback
Outcome data improves future recommendations.
1 operating angles mapped
Operational Depth
Technologies
Technologies commonly used in Oncology Eligibility Criteria Matcher implementations:
Real-World Use Cases
AI-assisted extraction of biomarker and TNM staging criteria for oncology trial matching
The AI is especially helpful at finding tricky cancer details like biomarker tests, biomarker results, and tumor staging in patient records, so staff can match patients to trials more accurately.
Criterion-level eligibility assessment for inclusion and exclusion rules
Instead of making one big yes-or-no decision immediately, the AI checks each trial rule one by one and shows whether the patient seems to meet or fail that specific rule.