Artificial intelligence in medicine. Conference on Artificial Intelligence in Medicine (2005- )最新文献

Approaches to guide empiric antimicrobial therapy are needed, especially in critically ill populations with prevalent antimicrobial resistance (AMR). While artificial intelligence shows promise in predicting AMR, scalable and generalizable prediction models are essential for broad clinical adoption. We utilized a publicly available clinical large language model (LLM), Gatortron, in comparison to traditional machine learning, to predict AMR and methicillin-resistant Staphylococcus aureus (MRSA)-specific patterns within a hospital-onset sepsis cohort using electronic health record (EHR) data available at time of illness onset. EHR data from approximately 150,000 hospitalizations with a documented bacterial infection at a large tertiary care healthcare system between 2010 and 2023 were examined. Among 2,019 eligible hospital-onset sepsis encounters, an AMR pathogen was identified in 911 (45%) and MRSA was isolated in 234 (26%). LLMs outperformed traditional models in predicting MRSA, achieving an AUC of 0.73 compared to 0.66 for the best traditional ML model, with superior F1 scores (0.43 vs. 0.16 for ML). Negative predictive value for MRSA prediction using LLM was at least 90% across majority of infection presentations. The LLM's superior prediction using a relatively simplified feature set demonstrates the potential of leveraging EHR data for early resistance prediction, though further refinement is needed to enhance sensitivity and clinical applicability.

Medication adherence is critical for the recovery of adolescents and young adults (AYAs) who have undergone hematopoietic cell transplantation. However, maintaining adherence is challenging for AYAs after hospital discharge, who experience both individual (e.g. physical and emotional symptoms) and interpersonal barriers (e.g., relational difficulties with their care partner, who is often involved in medication management). To optimize the effectiveness of a three-component digital intervention targeting both members of the dyad as well as their relationship, we propose a novel Multi-Agent Reinforcement Learning (MARL) approach to personalize the delivery of interventions. By incorporating the domain knowledge, the MARL framework, where each agent is responsible for the delivery of one intervention component, allows for faster learning compared with a flattened agent. Evaluation using a dyadic simulator environment, based on real clinical data, shows a significant improvement in medication adherence (approximately 3%) compared to purely random intervention delivery. The effectiveness of this approach will be further evaluated in an upcoming trial.

Electrocardiogram (ECG/EKG) analysis is a vital diagnostic tool for assessing heart conditions, extensively used in clinical applications such as patient monitoring, surgical support, and heart disease research. With the rising demand for automated EKG interpretation, particularly for disease diagnosis and waveform labeling, machine learning models have become essential. However, the scarcity of large, well-labeled EKG datasets poses a significant challenge for training EKG classification models in the supervised form. This has shifted the attention towards unsupervised model pre-training, which often outperforms pure supervised methods when only a limited number of labeled data is available. This study explores the adaptation of the contrastive representation learning framework for EKG classification. Traditional contrastive learning methods rely on data augmentations to create diverse views of the same sample, but these augmentations are domain-specific, difficult to design, and can unpredictably impact model performance across different tasks. In this work, we address these limitations by proposing a novel, augmentation-free approach that integrates seamlessly with existing contrastive frameworks by eliminating their dependence on augmentations and hence their potential drawbacks. We evaluate our approach on the PTB-XL [1] dataset, and demonstrate its benefits in the unsupervised model pre-training step. Our solution offers a promising pathway for enhancing cardiac disease diagnostics in data-constrained environments.

Traumatic experiences have the potential to give rise to post-traumatic stress disorder (PTSD), a debilitating psychiatric condition associated with impairments in both social and occupational functioning. There has been great interest in utilizing machine learning approaches to predict the development of PTSD in trauma patients from clinician assessment or survey-based psychological assessments. However, these assessments require a large number of questions, which is time consuming and not easy to administer. In this paper, we aim to predict PTSD development of patients 3 months post-trauma from multiple survey-based assessments taken within 2 weeks post-trauma. Our objective is to minimize the number of survey questions that patients need to answer while maintaining the prediction accuracy from the full surveys. We formulate this as a feature selection problem and consider 4 different feature selection approaches. We demonstrate that it is possible to achieve up to 72% accuracy for predicting the 3-month PTSD diagnosis from 10 survey questions using a mean decrease in impurity-based feature selector followed by a gradient boosting classifier.

The precise prediction of hypotension is vital for advancing preemptive patient care strategies. Traditional machine learning approaches, while instrumental in this field, are hampered by their dependence on structured historical data and manual feature extraction techniques. These methods often fall short of recognizing the intricate patterns present in physiological signals. Addressing this limitation, our study introduces an innovative application of deep learning technologies, utilizing a sophisticated end-to-end architecture grounded in XResNet. This architecture is further enhanced by the integration of contrastive learning and a value attention mechanism, specifically tailored to analyze arterial blood pressure (ABP) waveform signals. Our approach improves the performance of hypotension prediction over the existing state-of-theart ABP model [7]. This research represents a step towards optimizing patient care, embodying the next generation of AI-driven healthcare solutions. Through our findings, we demonstrate the promise of deep learning in overcoming the limitations of conventional prediction models, thereby offering an avenue for enhancing patient outcomes in clinical settings.

The availability of biomedical abstracts in online databases could improve health literacy and drive more informed choices. However, the technical language of these documents makes them inaccessible to healthcare consumers, causing disengagement, frustration and potential misuse. In this work we explore adapting foundation language models to the Plain Language Adaptation of Biomedical Abstracts benchmark. This task is challenging because it requires sentence-by-sentence simplifications, but entire abstracts must also be simplified cohesively. We present a sentence-wise autoregressive approach and report experiments with this technique in both zero-shot and fine-tuned settings, using both proprietary and open-source models. We also introduce a stochastic regularization technique to encourage recovery from source-copying during autoregressive inference. Our best-performing model achieves a 32 point increase in SARI and 6 point increase in BERTscore over the reported state-of-the-art. This also surpasses performance of recent open-domain and biomedical sentence simplification models on this task. Further, in manual evaluation, models achieve factual accuracy comparable to human-level, with simplicity close to that of humans. Abstracts simplified by these models could unlock a massive source of health information while retaining clear provenance for each statement to enhance trustworthiness.

Electrocardiogram (EKG/ECG) is a key diagnostic tool to assess patient's cardiac condition and is widely used in clinical applications such as patient monitoring, surgery support, and heart medicine research. With recent advances in machine learning (ML) technology there has been a growing interest in the development of models supporting automatic EKG interpretation and diagnosis based on past EKG data. The problem can be modeled as multi-label classification (MLC), where the objective is to learn a function that maps each EKG reading to a vector of diagnostic class labels reflecting the underlying patient condition at different levels of abstraction. In this paper, we propose and investigate an ML model that considers class-label dependency embedded in the hierarchical organization of EKG diagnoses to improve the EKG classification performance. Our model first transforms the EKG signals into a low-dimensional vector, and after that uses the vector to predict different class labels with the help of the conditional tree structured Bayesian network (CTBN) that is able to capture hierarchical dependencies among class variables. We evaluate our model on the publicly available PTB-XL dataset. Our experiments demonstrate that modeling of hierarchical dependencies among class variables improves the diagnostic model performance under multiple classification performance metrics as compared to classification models that predict each class label independently.

Kidney transplantation can significantly enhance living standards for people suffering from end-stage renal disease. A significant factor that affects graft survival time (the time until the transplant fails and the patient requires another transplant) for kidney transplantation is the compatibility of the Human Leukocyte Antigens (HLAs) between the donor and recipient. In this paper, we propose new biologically-relevant feature representations for incorporating HLA information into machine learning-based survival analysis algorithms. We evaluate our proposed HLA feature representations on a database of over 100,000 transplants and find that they improve prediction accuracy by about 1%, modest at the patient level but potentially significant at a societal level. Accurate prediction of survival times can improve transplant survival outcomes, enabling better allocation of donors to recipients and reducing the number of re-transplants due to graft failure with poorly matched donors.

Cancer patients experience many symptoms throughout their cancer treatment and sometimes suffer from lasting effects post-treatment. Patient-Reported Outcome (PRO) surveys provide a means for monitoring the patient's symptoms during and after treatment. Symptom cluster (SC) research seeks to understand these symptoms and their relationships to define new treatment and disease management methods to improve patient's quality of life. This paper introduces association rule mining (ARM) as a novel alternative for identifying symptom clusters. We compare the results to prior research and find that while some of the SCs are similar, ARM uncovers more nuanced relationships between symptoms such as anchor symptoms that serve as connections between interference and cancer-specific symptoms.