Searcharxiv⌕ Search

arXiv subjects

Brett K. Beaulieu-Jones

Publications and source records attributed to Brett K. Beaulieu-Jones.

11 recordsLinked to original sources
arXiv2026

Representation Before Training: A Practical Benchmark for Generative Medical Event Model Tokenization

Generative medical event models use tokenized sequences of patient timelines as input, but practical guidance on the many decisions around tokenization is limited. We benchmark quantization granularity, reference-range anchoring, code--value fusion, numeric and temporal encodings, and native versus harmonized event representations from an expert-mapped common data model. Using both Llama and Qwen architectures, 156 models were trained on full hospitalizations from three initialization seeds, with each configuration following a shared training recipe for up to five epochs. We evaluated learned representations from the first 24 hours of hospitalization with linear probes to predict binary and continuous outcomes during hours 24-48. Fused tokens pairing codes with value deciles increased performance across all eight outcome families relative to the equivalent unfused tokenized input with area under the receiver operating characteristic curve (AUROC) gains of $+0.002$ to $+0.033$ and Spearman correlation gains of $+0.025$ to $+0.114$. Neither anchoring value bins to reference ranges nor increasing quantization granularity consistently improved performance, while xVal variants underperformed both discrete and soft encodings. Alternatives to explicit time tokens, such as event-order and admission-relative rotary position embeddings (RoPE), yielded higher family-mean point estimates across all eight families while reducing input length. When evaluating native input against input mapped to the Common Longitudinal Intensive Care Unit Data Format (CLIF), the CLIF full-hospitalization training sequences contained 28.6% as many tokens as the native sequences and improved performance across six of eight outcome families. These findings show that tokenization and event encoding are consequential design choices when learning patient representations for downstream classification and regression tasks.

cs.LG↗
arXiv2026

Federated generative event models for tokenized electronic health records

Electronic health record foundation models are limited by institutionally siloed data and substantial performance degradation under cross-site transfer. We evaluated federated training of tokenized generative event models (GEMs) across 122,251 intensive care hospitalizations from three independent health systems harmonized to the Common Longitudinal ICU Data Format. Models were assessed on 12 post-24-hour clinical prediction tasks using within-site, cross-site, centralized, and federated training configurations. GEMs achieved the highest mean within-site and cross-site ROC-AUC and were substantially more transportable than conventional supervised models: their average cross-site penalties were 0.025 ROC-AUC and 0.027 PR-AUC, compared with 0.079 and 0.089 for LightGBM. Federated Learning (FedAvg and FedAvgM) approached the performance of centralized GEM training, with most gains obtained within 5-10 communication rounds. However, centralized multi-site training provided only modest improvements over complete local training. Multi-site models were most useful when local training data were limited, with their advantage narrowing as institutional data accumulated. These findings show that federated GEM training is technically feasible and preserves most centralized performance, but that the main open challenge is learning transportable representations to translate larger, but heterogeneous data from multiple health systems into a reliable target-site benefit.

cs.LG↗
arXiv2026

Efficient Generative Prediction for EHR Foundation Models: The SCOPE and REACH Estimators

Generative foundation models trained on tokenized electronic health record (EHR) timelines show promise for clinical outcome prediction via Monte Carlo sampling of simulated future trajectories. However, this approach suffers from three coupled limitations: sparse estimate distributions that poorly differentiate patient risk levels, extreme computational cost, and high sampling variance. We propose two new estimators that leverage next-token probability distributions underutilized by standard Monte Carlo: the Sum of Conditional Outcome Probability Estimator (SCOPE) and Risk Estimation from Anticipated Conditional Hazards (REACH). We prove both are unbiased, that REACH guarantees variance reduction over Monte Carlo for any model and outcome, and that REACH is a Rao-Blackwellization of any naive importance sampling scheme that preserves the non-outcome token distribution. Empirically, across $11$ clinically important outcomes in MIMIC-IV and the UChicago health system, SCOPE and REACH match $100$-sample Monte Carlo accuracy with median token reductions of $2.5\times$ to $3.4\times$ and reductions exceeding $80\times$ for the rarest outcomes, with calibration preserved throughout. Because SCOPE reuses a single sampled pool across an arbitrary number of outcomes at no marginal generation cost while REACH provides a per-task variance guarantee, the two estimators are complementary in deployment and together meaningfully reduce the inference budget required for generative EHR foundation models, particularly for rare, high-impact outcomes in healthcare.

stat.ML↗
arXiv2025

Quantifying surprise in clinical care: Detecting highly informative events in electronic health records with foundation models

We present a foundation model-derived method to identify highly informative tokens and events in electronic health records. Our approach considers incoming data in the entire context of a patient's hospitalization and so can flag anomalous events that rule-based approaches would consider within a normal range. We demonstrate that the events our model flags are significant for predicting downstream patient outcomes and that a fraction of events identified as carrying little information can safely be dropped. Additionally, we show how informativeness can help interpret the predictions of prognostic models trained on foundation model-derived representations.

cs.LG↗
arXiv2025

Foundation models for electronic health records: representation dynamics and transferability

Foundation models (FMs) trained on electronic health records (EHRs) have shown strong performance on a range of clinical prediction tasks. However, adapting these models to local health systems remains challenging due to limited data availability and resource constraints. In this study, we investigated what these models learn and evaluated the transferability of an FM trained on MIMIC-IV to an institutional EHR dataset at the University of Chicago Medical Center. We assessed their ability to identify outlier patients and examined representation-space patient trajectories in relation to future clinical outcomes. We also evaluated the performance of supervised fine-tuned classifiers on both source and target datasets. Our findings offer insights into the adaptability of FMs across different healthcare systems, highlight considerations for their effective implementation, and provide an empirical analysis of the underlying factors that contribute to their predictive performance.

cs.LG↗
arXiv2022

Harnessing electronic health records for real-world evidence

While randomized controlled trials (RCTs) are the gold-standard for establishing the efficacy and safety of a medical treatment, real-world evidence (RWE) generated from real-world data (RWD) has been vital in post-approval monitoring and is being promoted for the regulatory process of experimental therapies. An emerging source of RWD is electronic health records (EHRs), which contain detailed information on patient care in both structured (e. g., diagnosis codes) and unstructured (e. g., clinical notes, images) form. Despite the granularity of the data available in EHRs, critical variables required to reliably assess the relationship between a treatment and clinical outcome can be challenging to extract. We provide an integrated data curation and modeling pipeline leveraging recent advances in natural language processing, computational phenotyping, modeling techniques with noisy data to address this fundamental challenge and accelerate the reliable use of EHRs for RWE, as well as the creation of digital twins. The proposed pipeline is highly automated for the task and includes guidance for deployment. Examples are also drawn from existing literature on EHR emulation of RCT and accompanied by our own studies with Mass General Brigham (MGB) EHR.

stat.AP↗
arXiv2020

ML4H Abstract Track 2019

A collection of the accepted abstracts for the Machine Learning for Health (ML4H) workshop at NeurIPS 2019. This index is not complete, as some accepted abstracts chose to opt-out of inclusion.

cs.LG↗
arXiv2018

Privacy-Preserving Distributed Deep Learning for Clinical Data

Deep learning with medical data often requires larger samples sizes than are available at single providers. While data sharing among institutions is desirable to train more accurate and sophisticated models, it can lead to severe privacy concerns due the sensitive nature of the data. This problem has motivated a number of studies on distributed training of neural networks that do not require direct sharing of the training data. However, simple distributed training does not offer provable privacy guarantees to satisfy technical safe standards and may reveal information about the underlying patients. We present a method to train neural networks for clinical data in a distributed fashion under differential privacy. We demonstrate these methods on two datasets that include information from multiple independent sites, the eICU collaborative Research Database and The Cancer Genome Atlas.

cs.LG↗
arXiv2018

Learning Contextual Hierarchical Structure of Medical Concepts with Poincairé Embeddings to Clarify Phenotypes

Biomedical association studies are increasingly done using clinical concepts, and in particular diagnostic codes from clinical data repositories as phenotypes. Clinical concepts can be represented in a meaningful, vector space using word embedding models. These embeddings allow for comparison between clinical concepts or for straightforward input to machine learning models. Using traditional approaches, good representations require high dimensionality, making downstream tasks such as visualization more difficult. We applied Poincaré embeddings in a 2-dimensional hyperbolic space to a large-scale administrative claims database and show performance comparable to 100-dimensional embeddings in a euclidean space. We then examine disease relationships under different disease contexts to better understand potential phenotypes.

q-bio.QM↗
arXiv2017

Machine Learning for Structured Clinical Data

Research is a tertiary priority in the EHR, where the priorities are patient care and billing. Because of this, the data is not standardized or formatted in a manner easily adapted to machine learning approaches. Data may be missing for a large variety of reasons ranging from individual input styles to differences in clinical decision making, for example, which lab tests to issue. Few patients are annotated at a research quality, limiting sample size and presenting a moving gold standard. Patient progression over time is key to understanding many diseases but many machine learning algorithms require a snapshot, at a single time point, to create a usable vector form. Furthermore, algorithms that produce black box results do not provide the interpretability required for clinical adoption. This chapter discusses these challenges and others in applying machine learning techniques to the structured EHR (i.e. Patient Demographics, Family History, Medication Information, Vital Signs, Laboratory Tests, Genetic Testing). It does not cover feature extraction from additional sources such as imaging data or free text patient notes but the approaches discussed can include features extracted from these sources.

cs.LG↗