FRESH: Information-Geometric Calibration of Patient-Level Models to Aggregate Evidence
Abstract
This note introduces FRESH (Fusion of Recent Evidence and Subject Histories), a method for incorporating population-level summary results -- published clinical trials, registry summaries, prior natural-history studies, and peer-reviewed indirect comparisons -- into predictive models trained on patient-level data. This method provides a principled means of combining both patient-level and aggregate-level data types into a unified data-efficient model for clinical decision making. FRESH assumes access to a generative model trained on patient-level data sources (e.g. clinical trial or real-world data). The method produces patient-level predictions from a re-calibrated model that matches a set of specified aggregate statistics for a target population. This can be understood as a patient-level recapitulation of the aggregate source -- with the key property that the recalibration is a minimal perturbation of the original joint distribution in a specific information-geometric sense. The resulting samples can be analyzed directly or combined into a post-training procedure to update the original generative model. This approach enables several applications where rigorously incorporating patient-level data with summary information is valuable, including (i) contextualizing single-arm trial results with respect to recent standard-of-care, (ii) clinical-trial simulations for design and probability-of-technical-success estimation, and (iii) comparative-effectiveness analyses of on-market therapies.
Related
- Aggregate Calibration Can Mask Stress-State Undercoverage: Bounded-Memory Forecast Intervals under Recurrent Financial Stress
- Dynamic Mission Risk Management via Adaptive Information Acquisition and Online Uncertainty Calibration
- Maximum Entropy Probability Distributions on Spheres with Fixed Mean Busemann Function and Holomorphic-Information-Geometric Model of Cognition
- Does Calibration Pick the Right Model? Likelihood Evidence from Income Risk
- Information-Geometric Blind Spots of Conformal Test Martingales
BibTeX
@misc{fuller2026fresh,
title = {{FRESH: Information-Geometric Calibration of Patient-Level Models to Aggregate Evidence}},
author = {Franklin Fuller and Daniele Bertolini and Samantha Liang and Jason Christopher and Aaron M. Smith},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2605.16246},
eprint = {2605.16246},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2605.16246},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - FRESH: Information-Geometric Calibration of Patient-Level Models to Aggregate Evidence
AU - Franklin Fuller
AU - Daniele Bertolini
AU - Samantha Liang
AU - Jason Christopher
AU - Aaron M. Smith
PY - 2026
DA - 2026/10/21
PB - arXiv
DO - 10.48550/arxiv.2605.16246
UR - https://arxiv.org/abs/2605.16246
AB - This note introduces FRESH (Fusion of Recent Evidence and Subject Histories), a method for incorporating population-level summary results -- published clinical trials, registry summaries, prior natural-history studies, and peer-reviewed indirect comparisons -- into predictive models trained on patient-level data. This method provides a principled means of combining both patient-level and aggregate-level data types into a unified data-efficient model for clinical decision making. FRESH assumes access to a generative model trained on patient-level data sources (e.g. clinical trial or real-world data). The method produces patient-level predictions from a re-calibrated model that matches a set of specified aggregate statistics for a target population. This can be understood as a patient-level recapitulation of the aggregate source -- with the key property that the recalibration is a minimal perturbation of the original joint distribution in a specific information-geometric sense. The resulting samples can be analyzed directly or combined into a post-training procedure to update the original generative model. This approach enables several applications where rigorously incorporating patient-level data with summary information is valuable, including (i) contextualizing single-arm trial results with respect to recent standard-of-care, (ii) clinical-trial simulations for design and probability-of-technical-success estimation, and (iii) comparative-effectiveness analyses of on-market therapies.
ER -