Logo image
AgentQC: Policy-Constrained Agentic Assessment of Task-Aware Reliability in Medical Imaging Datasets
 

AgentQC: Policy-Constrained Agentic Assessment of Task-Aware Reliability in Medical Imaging Datasets

Gustavo Andrade-Miranda, Yiheng Cao, Tiphain Collet Pedro J Soto Vega
MICCAI 2026-29th International Conference on Medical Image Computing and Computer Assisted Intervention/2nd Agentic AI for Medicine Workshop (Strasbourg, France, 27/09/2026–01/10/2026)
2026
Pancreas CT Task-aware reliability Quality control Medical imaging Dataset curation Agentic AI

Medical imaging datasets are often curated using global quality-control rules, although case eligibility depends on the predefined requirements of the intended task. AgentQC is a static artifact-grounded agentic framework for task-aware medical dataset curation that separates deterministic evidence extraction, calibrated comparison, controlled reasoning, medical critique, review routing, deterministic final policy, and post-hoc evaluation into typed artifacts with explicit provenance. Reasoning and critique are explanatory and nonbinding; final curation actions remain controlled by deterministic evidence, routing, and policy logic. We instantiate AgentQC for 1,000 pancreas CT cases with visible-pancreas, pancreas-lesion, and lesion-subregion curation requirements. Internal evaluation across task profiles identified 350, 429, and 226 changes in final actions between profile pairs. Nevertheless, lesionpresent, confirmed-absent, and unresolved-evidence categories remained consistent across configurations. All cases contained complete reasoning and critique, and deterministic and calibrated modes produced identical final actions. These findings support artifact-backed compliance behavior under configured task requirements, not external medical correctness, expert agreement, or downstream model utility.

(1)

url
Find in HAL
1
Logo image