EviGround-CT: Sentence-to-Volume Grounding for Verifying LLM-Generated Chest CT Reports

A new method called EviGround-CT raises evidence-supported precision for LLM-generated chest CT reports from 0.263 to 0.365 across 137 extracted in-scope positive findings, according to an arXiv paper (2610.33038v1). The approach uses a deterministic extractor to identify positive finding sentences and a 3D grounding model to predict an evidence mask for each sentence, dropping the unsupported positive finding rate from 0.737 to 0.635 on a held-out 100-case ReXGroundingCT subset. The paired case-level bootstrap ESP difference was +0.102 (95% confidence interval: [0.028, 0.181]), with predicted masks linking individual report statements to spatial evidence inspectable alongside the CT scan.

arXiv:2610.33038v1 Announce Type: new Abstract: Chest computed tomography CT reports generated by large language models LLMs can describe positive findings that lack localized evidence in the scan. We propose EviGround-CT to verify these findings after report generation. A deterministic extractor identifies positive finding sentences, and a 3D grounding model predicts an evidence mask for each sentence. Using the full sentence retains anatomical location, laterality, and descriptive modifiers in the grounding query. We evaluate report generation on CT-RATE, and grounding and verification on a held-out 100-case ReXGroundingCT subset. Using the available annotations, sentence-grounded verification raises evidence-supported precision ESP from 0.263 to 0.365 across 137 extracted in-scope positives, compared with category conditioning. The corresponding unsupported positive finding rate falls from 0.737 to 0.635. The paired case-level bootstrap ESP difference is +0.102 95% confidence interval: 0.028, 0.181 . The predicted masks associate individual report statements with spatial evidence that can be inspected alongside the CT.