Integrating multi-modal clinical data, such as electronic health records (EHR) and chest X-ray images (CXR), is particularly beneficial for clinical prediction tasks. However, in a temporal setting, multi-modal data are often inherently asynchronous.

Objective: Large language models (LLMs) are attracting increasing interest in healthcare. This commentary evaluates the potential of LLMs to improve clinical prediction models (CPMs) for diagnostic and prognostic tasks, with a focus on their ability to process longitudinal electronic health record (EHR) data. Findings: LLMs show promise in handling multimodal and longitudinal EHR data and can support multi-outcome predictions for diverse health conditions. However, methodological, validation, infrastructural, and regulatory chal- lenges remain. These include inadequate methods for time-to-event modelling, poor calibration of predictions, limited external validation, and bias affecting underrepresented groups. High infrastructure costs and the absence of clear regulatory frameworks further prevent adoption. Implications: Further work and interdisciplinary collaboration are needed to support equitable and effective integra- tion into the clinical prediction. Developing temporally aware, fair, and explainable models should be a priority focus for transforming clinical prediction workflow.

Integrating multi-modal clinical data, such as electronic health records (EHR) and chest X-ray images (CXR), is particularly beneficial for clinical prediction tasks. However, in a temporal setting, multi-modal data are often inherently asynchronous. EHR can be continuously collected but CXR is generally taken with a much longer interval due to its high cost and radiation dose. When clinical prediction is needed, the last available CXR image might have been outdated, leading to suboptimal predictions. To address this challenge, we propose DDL-CXR, a method that dynamically generates an up-to-date latent representation of the individualized CXR images. Our approach leverages latent diffusion models for patient-specific generation strategically conditioned on a previous CXR image and EHR time series, providing information regarding anatomical structures and disease progressions, respectively. In this way, the interaction across modalities could be better captured by the latent CXR generation process, ultimately improving the prediction performance. Experiments using MIMIC datasets show that the proposed model could effectively address asynchronicity in multimodal fusion and consistently outperform existing methods.

We present Clinical Prediction with Large Language Models (CPLLM), a method that involves fine-tuning a pre-trained Large Language Model (LLM) for clinical disease prediction. We utilized quantization and fine-tuned the LLM using prompts, with the task of predicting whether patients will be diagnosed with a target disease during their next visit or in the subsequent diagnosis, leveraging their historical diagnosis records. We compared our results versus various baselines, including Logistic Regression, RETAIN, and Med-BERT, which is the current state-of-the-art model for disease prediction using structured EHR data. Our experiments have shown that CPLLM surpasses all the tested models in terms of both PR-AUC and ROC-AUC metrics, displaying noteworthy enhancements compared to the baseline models.

Clinical prediction is an essential task in the healthcare industry. However, the recent success of transformers, on which large language models are built, has not been extended to this domain. In this research, we explore the use of transformers and language models in prognostic prediction for immunotherapy using real-world patients' clinical data and molecular profiles. This paper investigates the potential of transformers to improve clinical prediction compared to conventional machine learning approaches and addresses the challenge of few-shot learning in predicting rare disease areas. The study benchmarks the efficacy of baselines and language models on prognostic prediction across multiple cancer types and investigates the impact of different pretrained language models under few-shot regimes. The results demonstrate significant improvements in accuracy and highlight the potential of NLP in clinical research to improve early detection and intervention for different diseases.

Cardiovascular Disease has long been the number one cause of death in the U.S. and some of the stats are startling: an American will have a heart attack approximately every 40 seconds for a total of 805,000 every year, At the same time, mortality and morbidity rates of CVD are increasing year by year, especially in developing regions. Studies have shown that approximately 80% of CVD-related deaths occur in low- and middle-income countries. Besides, these deaths occur at a younger age than in high-income countries. CVD represents a significant economic cost for society, around $351.2 billion in the US, chronically affecting patients' quality of life. The EU has estimated that the overall yearly cost amounts to €210 billion, allocating around 53% to healthcare costs (€111 billion), with 26% related to productivity losses (€54 billion), and the remaining 21% (€45 billion) to the informal care of people with CVD (European Cardiovascular Disease Statistics 2017).

Now, Johns Hopkins Medicine researchers have developed an advanced machine-learning system that can accurately predict how a patient's bout with COVID-19 will go, and relay its findings back to the clinician in an easily understandable form. The new prognostic tool, known as the Severe COVID-19 Adaptive Risk Predictor (SCARP), can help define the one-day and seven-day risk of a patient hospitalized with COVID-19 developing a more severe form of the disease or dying from it. SCARP asks for a minimal amount of input to give an accurate prediction, making it fast, simple to use and reliable for basing treatment and care decisions. The new tool is described in a paper first posted online March 2 in the Annals of Internal Medicine. "SCARP was designed to provide clinicians with a predictive tool that is interactive and adaptive, enabling real-time clinical variables to be entered at a patient's bedside," says Matthew Robinson, M.D., assistant professor of medicine at the Johns Hopkins University School of Medicine and senior author of the paper.

An advanced machine-learning system can accurately predict if a patient's bout with COVID-19 will become severe or fatal and relay its findings to clinicians. Clinicians often learn how to recognize patterns in COVID-19 cases after they treat many patients with it. Machine-learning systems promise to enhance that ability, recognizing more complex patterns in large numbers of people with COVID-19 and using that insight to predict the course of an individual patient's case. However, physicians sworn to "do no harm" may be reluctant to base treatment and care strategies for their most seriously ill patients on difficult-to-use or hard-to-interpret machine-learning algorithms. The new system offers findings in an easily understandable form.

You are free to share this article under the Attribution 4.0 International license. An advanced machine-learning system can accurately predict if a patient's bout with COVID-19 will become severe or fatal and relay its findings to clinicians. Clinicians often learn how to recognize patterns in COVID-19 cases after they treat many patients with it. Machine-learning systems promise to enhance that ability, recognizing more complex patterns in large numbers of people with COVID-19 and using that insight to predict the course of an individual patient's case. However, physicians sworn to "do no harm" may be reluctant to base treatment and care strategies for their most seriously ill patients on difficult-to-use or hard-to-interpret machine-learning algorithms.