Key result
PRiSM nomograms preserve ~90% of source model discrimination for 1-year mortality after heart transplantation.
Why the study?
Does the PRiSM framework maintain the predictive performance of black-box models for 1-year mortality in heart transplant recipients when converted to transparent nomograms?
Observational (n=50,356)
Yes
Does the PRiSM framework maintain the predictive performance of black-box models for 1-year mortality in heart transplant recipients when converted to transparent nomograms?
The PRiSM framework successfully translates complex black-box clinical prediction models into transparent, auditable nomograms without clinically significant loss of predictive performance for post-transplant mortality.
Answers are grounded in this paper and cite their sources.
May enable transparent, auditable 1-year mortality prediction after heart transplant; leaves open prospective multicenter validation before clinical adoption.
PRiSM (Partial Responses in Structured Models) is a model-agnostic framework that converts a trained probabilistic binary classifier for tabular data into a standalone nomogram that can be audited term by term. PRiSM takes the shape of each nonlinear effect and interaction from the source model, not merely which variables mattered, and lets the observed outcome select and weight them. The result is an additive model that replaces the original classifier as the deployed predictor. Because the nomogram is additive on the log-odds scale, each variable's contribution can be read off and summed by hand. PRiSM is a collaboration between the Artificial Intelligence and Bioinformatics in Cardiothoracic Sciences Research Unit at Lund University, Sweden, and the Artificial Intelligence and Digital Technologies Research Institute at Liverpool John Moores University, UK. The framework is authored by: Henry Pigot (1), Paulo J. G. Lisboa (2), Sandra Ortega-Martorell (2), Ivan Olier (2), Joseph Mahon (2), Johan Nilsson (1,3) (1) Department of Translational Medicine, Artificial Intelligence and Bioinformatics in Cardiothoracic Sciences Research Unit, Lund University, Lund, Sweden (2) Artificial Intelligence and Digital Technologies Research Institute, Liverpool John Moores University, Liverpool, UK (3) Department of Thoracic and Vascular Surgery, Skåne University Hospital, Lund, Sweden This record archives the v0.1.1 source release. Software authorship: the code archived here was written by Henry Pigot, Paulo J. G. Lisboa, and Johan Nilsson, and is released under the BSD 3-Clause License. If you use PRiSM in published work, please cite: Pigot, H., Lisboa, P. J. G., Ortega-Martorell, S., Olier, I., Mahon, J., & Nilsson, J. (2026). Translation of black-box clinical prediction models into standalone transparent nomograms: Temporal external validation in heart transplantation. arXiv. https://doi.org/10.48550/arXiv.2609.07610 Source and releases: https://github.com/AIBCTS/PRiSM Package: https://pypi.org/project/prism-iml/ Install: pip install prism-iml
No takes yet. Share an insight, caveat, or question.
Pigot et al. (2026) conducted an observational in Heart transplantation (n=50,356). PRiSM nomograms vs. Source black-box models was evaluated on Noninferiority in temporal external-validation AUROC (preservation of at least 90% of the source model's discrimination above chance). Baseline PRiSM nomograms derived from 5 source models met a prespecified noninferiority criterion for discrimination of 1-year mortality after heart transplantation in temporal external validation.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: