Search papers, labs, and topics across Lattice.
This paper introduces a novel training-free best-of-N sampling method for generating chest X-ray reports that incorporates the longitudinal context of prior exams. By utilizing a transition-aware approach, the method encodes changes between prior and current reports into directional vectors, allowing for more relevant and contextually informed report generation. The results demonstrate significant improvements in report quality, particularly in the Impression section, outperforming traditional random selection methods.
Transition-aware sampling can dramatically enhance the relevance of chest X-ray reports by leveraging longitudinal patient data, yielding superior results in clinical settings.
In longitudinal clinical practice, every chest X-ray is read in the context of the patients prior exam, and much of what the radiologist communicates is the change from one visit to the next. To the best of our knowledge, we present the first training-free best-of-N sampling scheme for pre-trained chest X-ray report generators that is explicitly aware of this longitudinal prior to current transition. We call it transition-aware best-of-N sampling, each report is split into sentences and embedded into an unordered set in Rd; each (prior, current) pair is reduced to a fixed-dim directional vector via a set-to-set distance designed to encode the change between the two sets; and candidates are scored by cosine distance from their candidate transition vector to a cached bank of ground-truth training transition vectors, aggregated as min or kNN. We instantiate the framework with four directional set distances (mean-shift, novelty residual, directed-Hausdorff anchor, and cost-weighted optimal transport) and evaluate on a multi-visit AP-PA cohort, running inference under three prompts on three vision-language generators. Transition-aware best-of-N outperforms random selection across the board, with the largest relative gains on the Impression section.
