PRISM: A Transformer-based Language Model of Structured Clinical Event Data

TL;DR

PRISM is a transformer-based model that effectively captures complex sequential clinical event data, improving prediction of diagnostic pathways and supporting clinical decision-making.

Contribution

It introduces a novel transformer architecture tailored for structured clinical sequences, enabling realistic modeling of patient diagnostic trajectories.

Findings

Significant improvement over baselines in next-token prediction

Sequences reflect realistic diagnostic and laboratory progressions

Demonstrates potential for clinical decision support applications

Abstract

We introduce PRISM (Predictive Reasoning in Sequential Medicine), a transformer-based architecture designed to model the sequential progression of clinical decision-making processes. Unlike traditional approaches that rely on isolated diagnostic classification, PRISM frames clinical trajectories as tokenized sequences of events - including diagnostic tests, laboratory results, and diagnoses - and learns to predict the most probable next steps in the patient diagnostic journey. Leveraging a large custom clinical vocabulary and an autoregressive training objective, PRISM demonstrates the ability to capture complex dependencies across longitudinal patient timelines. Experimental results show substantial improvements over random baselines in next-token prediction tasks, with generated sequences reflecting realistic diagnostic pathways, laboratory result progressions, and clinician ordering behaviors. These findings highlight the feasibility of applying generative language modeling techniques to structured medical event data, enabling applications in clinical decision support, simulation, and education. PRISM establishes a foundation for future advancements in sequence-based healthcare modeling, bridging the gap between machine learning architectures and real-world diagnostic reasoning.