Contrastive Cross-Modal Learning for Infusing Chest X-ray Knowledge into ECGs

Modern diagnostic workflows are increasingly multimodal, integrating diverse data sources such as medical images, structured records, and physiological time series. Among these, electrocardiograms (ECGs) and chest X-rays (CXRs) are two of the most widely used modalities for cardiac assessment. While CXRs provide rich diagnostic information, ECGs are more accessible and can support scalable early warning systems. In this work, we propose CroMoTEX, a novel contrastive learning-based framework that leverages chest X-rays during training to learn clinically informative ECG representations for multiple cardiac-related pathologies: cardiomegaly, pleural effusion, and edema. Our method aligns ECG and CXR representations using a novel supervised cross-modal contrastive objective with adaptive hard negative weighting, enabling robust and task-relevant feature learning. At test time, CroMoTEX relies solely on ECG input, allowing scalable deployment in real-world settings where CXRs may be unavailable. Evaluated on the large-scale MIMIC-IV-ECG and MIMIC-CXR datasets, CroMoTEX outperforms baselines across all three pathologies, achieving up to 78.31 AUROC on edema. Our code is available at github.com/vineetpmoorty/cromotex.

DeVisE: Behavioral Testing of Medical Large Language Models

Large language models (LLMs) are increasingly used in clinical decision support, yet current evaluation methods often fail to distinguish genuine medical reasoning from superficial patterns. We introduce DeVisE (Demographics and Vital signs Evaluation), a behavioral testing framework for probing fine-grained clinical understanding. We construct a dataset of ICU discharge notes from MIMIC-IV, generating both raw (real-world) and template-based (synthetic) versions with controlled single-variable counterfactuals targeting demographic (age, gender, ethnicity) and vital sign attributes. We evaluate five LLMs spanning general-purpose and medically fine-tuned variants, under both zero-shot and fine-tuned settings. We assess model behavior via (1) input-level sensitivity - how counterfactuals alter the likelihood of a note; and (2) downstream reasoning - how they affect predicted hospital length-of-stay. Our results show that zero-shot models exhibit more coherent counterfactual reasoning patterns, while fine-tuned models tend to be more stable yet less responsive to clinically meaningful changes. Notably, demographic factors subtly but consistently influence outputs, emphasizing the importance of fairness-aware evaluation. This work highlights the utility of behavioral testing in exposing the reasoning strategies of clinical LLMs and informing the design of safer, more transparent medical AI systems.

Path-specific effects for pulse-oximetry guided decisions in critical care

Identifying and measuring biases associated with sensitive attributes is a crucial consideration in healthcare to prevent treatment disparities. One prominent issue is inaccurate pulse oximeter readings, which tend to overestimate oxygen saturation for dark-skinned patients and misrepresent supplemental oxygen needs. Most existing research has revealed statistical disparities linking device errors to patient outcomes in intensive care units (ICUs) without causal formalization. In contrast, this study causally investigates how racial discrepancies in oximetry measurements affect invasive ventilation in ICU settings. We employ a causal inference-based approach using path-specific effects to isolate the impact of bias by race on clinical decision-making. To estimate these effects, we leverage a doubly robust estimator, propose its self-normalized variant for improved sample efficiency, and provide novel finite-sample guarantees. Our methodology is validated on semi-synthetic data and applied to two large real-world health datasets: MIMIC-IV and eICU. Contrary to prior work, our analysis reveals minimal impact of racial discrepancies on invasive ventilation rates. However, path-specific effects mediated by oxygen saturation disparity are more pronounced on ventilation duration, and the severity differs by dataset. Our work provides a novel and practical pipeline for investigating potential disparities in the ICU and, more crucially, highlights the necessity of causal methods to robustly assess fairness in decision-making.

UdonCare: Hierarchy Pruning for Unseen Domain Discovery in Predictive Healthcare

Domain generalization has become a critical challenge in clinical prediction, where patient cohorts often exhibit shifting data distributions that degrade model performance. Typical domain generalization approaches struggle in real-world healthcare settings for two main reasons: (1) patient-specific domain labels are typically unavailable, making domain discovery especially difficult; (2) purely data-driven approaches overlook key clinical insights, leading to a gap in medical knowledge integration. To address these problems, we leverage hierarchical medical ontologies like the ICD-9-CM hierarchy to group diseases into higher-level categories and discover more flexible latent domains. In this paper, we introduce UdonCare, a hierarchy-guided framework that iteratively prunes fine-grained domains, encodes these refined domains, and applies a Siamese-type inference mechanism to separate domain-related signals from patient-level features. Experimental results on clinical datasets (MIMIC-III and MIMIC-IV) show that the proposed model achieves higher performance compared to other domain generalization baselines when substantial domain gaps presents, highlighting the untapped potential of medical knowledge for enhancing domain generalization in practical healthcare applications.

Exploring Scaling Laws for EHR Foundation Models

The emergence of scaling laws has profoundly shaped the development of large language models (LLMs), enabling predictable performance gains through systematic increases in model size, dataset volume, and compute. Yet, these principles remain largely unexplored in the context of electronic health records (EHRs) -- a rich, sequential, and globally abundant data source that differs structurally from natural language. In this work, we present the first empirical investigation of scaling laws for EHR foundation models. By training transformer architectures on patient timeline data from the MIMIC-IV database across varying model sizes and compute budgets, we identify consistent scaling patterns, including parabolic IsoFLOPs curves and power-law relationships between compute, model parameters, data size, and clinical utility. These findings demonstrate that EHR models exhibit scaling behavior analogous to LLMs, offering predictive insights into resource-efficient training strategies. Our results lay the groundwork for developing powerful EHR foundation models capable of transforming clinical prediction tasks and advancing personalized healthcare.

Reinforcement Learning for Out-of-Distribution Reasoning in LLMs: An Empirical Study on Diagnosis-Related Group Coding

Diagnosis-Related Group (DRG) codes are essential for hospital reimbursement and operations but require labor-intensive assignment. Large Language Models (LLMs) struggle with DRG coding due to the out-of-distribution (OOD) nature of the task: pretraining corpora rarely contain private clinical or billing data. We introduce DRG-Sapphire, which uses large-scale reinforcement learning (RL) for automated DRG coding from clinical notes. Built on Qwen2.5-7B and trained with Group Relative Policy Optimization (GRPO) using rule-based rewards, DRG-Sapphire introduces a series of RL enhancements to address domain-specific challenges not seen in previous mathematical tasks. Our model achieves state-of-the-art accuracy on the MIMIC-IV benchmark and generates physician-validated reasoning for DRG assignments, significantly enhancing explainability. Our study further sheds light on broader challenges of applying RL to knowledge-intensive, OOD tasks. We observe that RL performance scales approximately linearly with the logarithm of the number of supervised fine-tuning (SFT) examples, suggesting that RL effectiveness is fundamentally constrained by the domain knowledge encoded in the base model. For OOD tasks like DRG coding, strong RL performance requires sufficient knowledge infusion prior to RL. Consequently, scaling SFT may be more effective and computationally efficient than scaling RL alone for such tasks.

Predicting Length of Stay in Neurological ICU Patients Using Classical Machine Learning and Neural Network Models: A Benchmark Study on MIMIC-IV

Intensive care unit (ICU) is a crucial hospital department that handles life-threatening cases. Nowadays machine learning (ML) is being leveraged in healthcare ubiquitously. In recent years, management of ICU became one of the most significant parts of the hospital functionality (largely but not only due to the worldwide COVID-19 pandemic). This study explores multiple ML approaches for predicting LOS in ICU specifically for the patients with neurological diseases based on the MIMIC-IV dataset. The evaluated models include classic ML algorithms (K-Nearest Neighbors, Random Forest, XGBoost and CatBoost) and Neural Networks (LSTM, BERT and Temporal Fusion Transformer). Given that LOS prediction is often framed as a classification task, this study categorizes LOS into three groups: less than two days, less than a week, and a week or more. As the first ML-based approach targeting LOS prediction for neurological disorder patients, this study does not aim to outperform existing methods but rather to assess their effectiveness in this specific context. The findings provide insights into the applicability of ML techniques for improving ICU resource management and patient care. According to the results, Random Forest model proved to outperform others on static, achieving an accuracy of 0.68, a precision of 0.68, a recall of 0.68, and F1-score of 0.67. While BERT model outperformed LSTM model on time-series data with an accuracy of 0.80, a precision of 0.80, a recall of 0.80 and F1-score 0.80.

PatientSim: A Persona-Driven Simulator for Realistic Doctor-Patient Interactions

Doctor-patient consultations require multi-turn, context-aware communication tailored to diverse patient personas. Training or evaluating doctor LLMs in such settings requires realistic patient interaction systems. However, existing simulators often fail to reflect the full range of personas seen in clinical practice. To address this, we introduce PatientSim, a patient simulator that generates realistic and diverse patient personas for clinical scenarios, grounded in medical expertise. PatientSim operates using: 1) clinical profiles, including symptoms and medical history, derived from real-world data in the MIMIC-ED and MIMIC-IV datasets, and 2) personas defined by four axes: personality, language proficiency, medical history recall level, and cognitive confusion level, resulting in 37 unique combinations. We evaluated eight LLMs for factual accuracy and persona consistency. The top-performing open-source model, Llama 3.3, was validated by four clinicians to confirm the robustness of our framework. As an open-source, customizable platform, PatientSim provides a reproducible and scalable solution that can be customized for specific training needs. Offering a privacy-compliant environment, it serves as a robust testbed for evaluating medical dialogue systems across diverse patient presentations and shows promise as an educational tool for healthcare.

Development of Interactive Nomograms for Predicting Short-Term Survival in ICU Patients with Aplastic Anemia

Aplastic anemia is a rare, life-threatening hematologic disorder characterized by pancytopenia and bone marrow failure. ICU admission in these patients often signals critical complications or disease progression, making early risk assessment crucial for clinical decision-making and resource allocation. In this study, we used the MIMIC-IV database to identify ICU patients diagnosed with aplastic anemia and extracted clinical features from five domains: demographics, synthetic indicators, laboratory results, comorbidities, and medications. Over 400 variables were reduced to seven key predictors through machine learning-based feature selection. Logistic regression and Cox regression models were constructed to predict 7-, 14-, and 28-day mortality, and their performance was evaluated using AUROC. External validation was conducted using the eICU Collaborative Research Database to assess model generalizability. Among 1,662 included patients, the logistic regression model demonstrated superior performance, with AUROC values of 0.8227, 0.8311, and 0.8298 for 7-, 14-, and 28-day mortality, respectively, compared to the Cox model. External validation yielded AUROCs of 0.7391, 0.7119, and 0.7093. Interactive nomograms were developed based on the logistic regression model to visually estimate individual patient risk. In conclusion, we identified a concise set of seven predictors, led by APS III, to build validated and generalizable nomograms that accurately estimate short-term mortality in ICU patients with aplastic anemia. These tools may aid clinicians in personalized risk stratification and decision-making at the point of care.

No Black Boxes: Interpretable and Interactable Predictive Healthcare with Knowledge-Enhanced Agentic Causal Discovery

Deep learning models trained on extensive Electronic Health Records (EHR) data have achieved high accuracy in diagnosis prediction, offering the potential to assist clinicians in decision-making and treatment planning. However, these models lack two crucial features that clinicians highly value: interpretability and interactivity. The ``black-box'' nature of these models makes it difficult for clinicians to understand the reasoning behind predictions, limiting their ability to make informed decisions. Additionally, the absence of interactive mechanisms prevents clinicians from incorporating their own knowledge and experience into the decision-making process. To address these limitations, we propose II-KEA, a knowledge-enhanced agent-driven causal discovery framework that integrates personalized knowledge databases and agentic LLMs. II-KEA enhances interpretability through explicit reasoning and causal analysis, while also improving interactivity by allowing clinicians to inject their knowledge and experience through customized knowledge bases and prompts. II-KEA is evaluated on both MIMIC-III and MIMIC-IV, demonstrating superior performance along with enhanced interpretability and interactivity, as evidenced by its strong results from extensive case studies.