DIST-RRG: DISTRIBUTIONAL CHANGE MODELING ON STATISTICAL MANIFOLDS FOR LONGITUDINAL RADIOLOGY REPORT GENERATION

Longitudinal radiology reports describe the current study in relation to earlier visits. Pointwise change representations retain estimated state but discard the uncertainty attached to each reading. We present Dist-RRG, a distributional framework that represents visits, between-visit changes, and prior reports as diagonal Gaussians. A distributional difference module retains both visits' variances; a trajectory-conditioned, Bhattacharyya-style dissimilarity scores priors; and moment-matched aggregation preserves within-prior uncertainty and between-prior disagreement. Aggregated log-variance modulates the decoder through FiLM. On Longitudinal MIMIC-CXR, micro-averaged CheXbert clinical-efficacy F1 is $0.509$ with one prior and $0.516$ with up to five priors ($0.515\pm 0.003$ over three training seeds). Under a matched protocol - same split, evaluation script, and averaging - the strongest re-trained open-source longitudinal baseline reaches $0.482$ CE-F1, below Dist-RRG (P1) with a paired interval that excludes zero. The evaluation focuses on whether the propagated second moment carries useful information: progression calibration has $15$-bin ECE $0.041$ versus $0.063$ without moment-matched aggregation; aggregated variance detects unreliable targets with AUROC $0.712$; and retaining the least uncertain $70\%$ raises CE-F1 to $0.553$. At a fixed checkpoint, permuting variance vectors across patients reduces New-transition F1 from $0.412$ to $0.389$. These analyses support a bounded conclusion: propagated variance carries patient-specific information about prediction difficulty. They do not establish diagnostic confidence or superiority to other uncertainty estimators.