Proceedings of machine learning research最新文献_第5页

What do LLMs value? An evaluation framework for revealing subjective trade-offs in assessment of glycemic control. 法学硕士看重什么？在血糖控制评估中揭示主观权衡的评估框架。

Proceedings of machine learning research Pub Date : 2025-12-01

Payal Chandak, Elizabeth Healey, Maria F Villa-Tamayo, Agatha F Scheideman, Mandy M Shao, Chiara Fabris, Kenneth D Mandl, Isaac Kohane, David C Klonoff

{"title":"What do LLMs value? An evaluation framework for revealing subjective trade-offs in assessment of glycemic control.","authors":"Payal Chandak, Elizabeth Healey, Maria F Villa-Tamayo, Agatha F Scheideman, Mandy M Shao, Chiara Fabris, Kenneth D Mandl, Isaac Kohane, David C Klonoff","doi":"","DOIUrl":"","url":null,"abstract":"Clinical decisions often require balancing conflicting priorities rather than simply selecting a single \"correct\" answer. We present an evaluation framework that probes the value judgments embedded in large language models (LLMs) by testing how they assess quality of glycemic control from continuous glucose monitoring (CGM) data. Using synthetic type 1 diabetes profiles, we asked five commercial LLMs to perform pairwise comparisons of CGM summary statistics and derived a percentile ranking for each profile. We then quantified alignment with two reference metrics: time in range (TIR) and the expert-derived Glycemia Risk Index (GRI), which was developed with clinician input regarding preferences across glycemic ranges. Across three insulin therapy modalities, newer models showed stronger correlation with GRI than older models, suggesting a generational shift toward expert consensus. However, a perturbation analysis revealed instances of disagreement around the weighting of mild hypoglycemia and mild hyperglycemia relative to the GRI. These results demonstrate that high average agreement with clinical metrics can mask clinically meaningful misalignments in how LLMs prioritize risks. Our proposed framework reveals how LLM outputs reflect competing priorities in clinical contexts.","PeriodicalId":74504,"journal":{"name":"Proceedings of machine learning research","volume":"297 ","pages":"136-151"},"PeriodicalIF":0.0,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13322355/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148371211","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Anatomy-Guided Surface Diffusion Model for Alzheimer's Disease Normative Modeling. 解剖引导的表面扩散模型用于阿尔茨海默病规范建模。

Proceedings of machine learning research Pub Date : 2025-12-01

Jianwei Zhang, Yonggang Shi

{"title":"Anatomy-Guided Surface Diffusion Model for Alzheimer's Disease Normative Modeling.","authors":"Jianwei Zhang, Yonggang Shi","doi":"","DOIUrl":"","url":null,"abstract":"Normative modeling has emerged as a pivotal approach for characterizing heterogeneity and individual variance in neurodegenerative diseases, notably Alzheimer's disease (AD). One of the challenges of cortical normative modeling is the anatomical structure mismatch due to folding pattern variability. Traditionally, registration is applied to address this issue and recently deep generative models are employed to generate anatomically aligned samples for analyzing disease progression; however, these models are predominantly applied to volume-based data, which often falls short in capturing intricate morphological changes on the brain cortex. As an alternative, surface-based analysis has been proven to be more sensitive in disease modeling such as AD. Yet, like volume-based data, it also suffers from the mismatch problem. To address these limitations, we propose a novel generative normative modeling framework by transferring the conditional diffusion generative model to the spherical domain. Furthermore, the proposed model generates normal feature map distributions by explicitly conditioning on individual anatomical segmentation to ensure better geometrical alignment which helps to reduce variance between subjects in normative analysis. We find that our model can generate samples that are better anatomically aligned than registered reference data and through ablation study and normative assessment experiments, the samples are able to better measure individual differences from the normal distribution and increase sensitivity in differentiating cognitively normal (CN), mild cognitive impairment (MCI), and Alzheimer's disease (AD) patients.","PeriodicalId":74504,"journal":{"name":"Proceedings of machine learning research","volume":"301 ","pages":"1866-1878"},"PeriodicalIF":0.0,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13197061/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148019224","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

PhysioJEPA: Joint Embedding Representations of Physiological Signals for Real Time Risk Estimation in the Intensive Care Unit. 用于重症监护病房实时风险评估的生理信号联合嵌入表示。

Proceedings of machine learning research Pub Date : 2025-12-01

Benjamin Fox, Dung Hoang, Joy Jiang, Pushkala Jayaraman, Ankit Parekh, Girish N Nadkarni, Ankit Sakhuja

{"title":"PhysioJEPA: Joint Embedding Representations of Physiological Signals for Real Time Risk Estimation in the Intensive Care Unit.","authors":"Benjamin Fox, Dung Hoang, Joy Jiang, Pushkala Jayaraman, Ankit Parekh, Girish N Nadkarni, Ankit Sakhuja","doi":"","DOIUrl":"","url":null,"abstract":"Self-supervised learning of multi-modal, high-frequency physiological signals is largely unexplored, despite its potential for critical care applications. We present PhysioJEPA, a Joint Embedding Predictive Architecture (JEPA) designed for multi-modal physiological signals from critical care bedside monitoring devices. PhysioJEPA learns representations from 30-minute segments of physiological signals from three channels: arterial blood pressure, electrocardiography lead II, and photoplethysmography. Trained on over 10.7 million minutes of data from 4,282 intensive care unit stays (N=2,631 patients) in the Medical Information Mart for Intensive Care-III (MIMIC-III) Waveform Database, the learned, frozen representations of PhysioJEPA can be used to estimate 5-minute risk of hypotension (AUROC = 0.83 [Confidence Interval or CI 0.83-0.84]) and shock index (AUROC = 0.95 [0.95-0.96]), with comparable performance to a self-supervised Patch Time Series Transformer framework (AUROC = 0.87 [0.86-0.87] and 0.96 [0.96-0.96]), better performance compared to another JEPA physiological signal model, ECG-JEPA (AUROC = 0.73 [0.72-74] and 0.92 [0.92-0.93]), and better performance compared to a supervised convolutional model (AUROC = 0.78 [0.78-0.78] and 0.95 [0.95-0.95]). Notably, it can generalize to an independent healthcare system (AUROC = 0.78 [0.78-0.78] and 0.92 [0.92-0.93]) better than all comparison models. These results suggest that self-supervised JEPA representation learning is a promising approach for multi-modal bedside monitoring signal data.","PeriodicalId":74504,"journal":{"name":"Proceedings of machine learning research","volume":"297 ","pages":"120-135"},"PeriodicalIF":0.0,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13152001/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147870662","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Beyond Diagnosis: Evaluating Multimodal LLMs for Pathology Localization in Chest Radiographs. 超越诊断：评估胸片病理定位的多模态llm。

Proceedings of machine learning research Pub Date : 2025-12-01

Advait Gosai, Arun Kavishwar, Stephanie L McNamara, Soujanya Samineni, Renato Umeton, Alexander Chowdhury, William Lotter

{"title":"Beyond Diagnosis: Evaluating Multimodal LLMs for Pathology Localization in Chest Radiographs.","authors":"Advait Gosai, Arun Kavishwar, Stephanie L McNamara, Soujanya Samineni, Renato Umeton, Alexander Chowdhury, William Lotter","doi":"","DOIUrl":"","url":null,"abstract":"Recent work has shown promising performance of frontier large language models (LLMs) and their multimodal counterparts in medical quizzes and diagnostic tasks, highlighting their potential for broad clinical utility given their accessible, general-purpose nature. However, beyond diagnosis, a fundamental aspect of medical image interpretation is the ability to localize pathological findings. Evaluating localization not only has clinical and educational relevance but also provides insight into a model's spatial understanding of anatomy and disease. Here, we systematically assess two general-purpose MLLMs (GPT-4 and GPT-5) and a domain-specific model (MedGemma) in their ability to localize pathologies on chest radiographs, using a prompting pipeline that overlays a spatial grid and elicits coordinate-based predictions. Averaged across nine pathologies in the CheXlocalize dataset, GPT-5 exhibited a localization accuracy of 49.7%, followed by GPT-4 (39.1%) and MedGemma (17.7%), all lower than a task-specific CNN baseline (59.9%) and a radiologist benchmark (80.1%). Despite modest performance, error analysis revealed that GPT-5's predictions were largely in anatomically plausible regions, just not always precisely localized. GPT-4 performed well on pathologies with fixed anatomical locations, but struggled with spatially variable findings and exhibited anatomically implausible predictions more frequently. MedGemma demonstrated the lowest performance on all pathologies, but showed improvements when provided examples through few shot prompting. Our findings highlight both the promise and limitations of current MLLMs in medical imaging and underscore the importance of integrating them with task-specific tools for reliable use.","PeriodicalId":74504,"journal":{"name":"Proceedings of machine learning research","volume":"297 ","pages":"572-587"},"PeriodicalIF":0.0,"publicationDate":"2025-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13225679/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148152455","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Self-Contemplating In-Context Learning Enhances T Cell Receptor Generation for Novel Epitopes. 自我思考上下文学习增强新表位的T细胞受体生成。

Proceedings of machine learning research Pub Date : 2025-09-01

Pengfei Zhang, Sonal Sujit Prabhu, Gloria Grama, Seojin Bang, Heewook Lee

{"title":"Self-Contemplating In-Context Learning Enhances T Cell Receptor Generation for Novel Epitopes.","authors":"Pengfei Zhang, Sonal Sujit Prabhu, Gloria Grama, Seojin Bang, Heewook Lee","doi":"","DOIUrl":"","url":null,"abstract":"Computational design of T cell receptors (TCRs) that bind to epitopes holds the potential to revolutionize targeted immunotherapy. However, computational design of TCRs for novel epitopes is challenging due to the scarcity of training data, and the absence of known cognate TCRs for novel epitopes. In this study, we aim to generate high-quality cognate TCRs particularly for novel epitopes with no known cognate TCRs, a problem that remains under-explored in the field. We propose to incorporate in-context learning, successfully used with large language models to perform new generative tasks, to the task of TCR generation for novel epitopes. By providing cognate TCRs as additional context, we enhance the model's ability to generate high-quality TCRs for novel epitopes. We first unlock the power of in-context learning by training a model to generate new TCRs based on both a target epitope and a small set of its cognate TCRs, so-called in-context training (ICT). We then self-generate its own TCR contexts based on a target epitope, as novel epitopes lack known binding TCRs, and use it as an inference prompt, referred to as self-contemplation prompting (SCP). Our experiments first demonstrate that aligning training and inference distribution by ICT is critical for effectively leveraging context TCRs. Subsequently, we show that providing context TCRs significantly improves TCR generation for novel epitopes. Furthermore, we show TCR generation using SCP-synthesized context TCRs achieves performance comparable to, and sometimes surpassing, ground-truth context TCRs, especially when combined with refined prompt selection based on binding affinity and authenticity metrics. We assess the designed sequences' binding probability and sequence authenticity using seven diverse computational models.","PeriodicalId":74504,"journal":{"name":"Proceedings of machine learning research","volume":"311 ","pages":"251-269"},"PeriodicalIF":0.0,"publicationDate":"2025-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13103298/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147791471","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

ProtoECGNet: Case-Based Interpretable Deep Learning for Multi-Label ECG Classification with Contrastive Learning. ProtoECGNet：基于案例的可解释深度学习的多标签心电分类对比学习。

Proceedings of machine learning research Pub Date : 2025-08-01

Sahil Sethi, David Chen, Thomas Statchen, Michael C Burkhart, Nipun Bhandari, Bashar Ramadan, Brett Beaulieu-Jones

{"title":"ProtoECGNet: Case-Based Interpretable Deep Learning for Multi-Label ECG Classification with Contrastive Learning.","authors":"Sahil Sethi, David Chen, Thomas Statchen, Michael C Burkhart, Nipun Bhandari, Bashar Ramadan, Brett Beaulieu-Jones","doi":"","DOIUrl":"","url":null,"abstract":"Deep learning-based electrocardiogram (ECG) classification has shown impressive performance but clinical adoption has been slowed by the lack of transparent and faithful explanations. Post hoc methods such as saliency maps may fail to reflect a model's true decision process. Prototype-based reasoning offers a more transparent alternative by grounding decisions in similarity to learned representations of real ECG segments-enabling faithful, case-based explanations. We introduce ProtoECGNet, a prototype-based deep learning model for interpretable, multi-label ECG classification. ProtoECGNet employs a structured, multi-branch architecture that reflects clinical interpretation workflows: it integrates a 1D CNN with global prototypes for rhythm classification, a 2D CNN with time-localized prototypes for morphology-based reasoning, and a 2D CNN with global prototypes for diffuse abnormalities. Each branch is trained with a prototype loss designed for multi-label learning, combining clustering, separation, diversity, and a novel contrastive loss that encourages appropriate separation between prototypes of unrelated classes while allowing clustering for frequently co-occurring diagnoses. We evaluate ProtoECGNet on all 71 labels from the PTB-XL dataset, demonstrating competitive performance relative to state-of-the-art black-box models while providing structured, case-based explanations. To assess prototype quality, we conduct a structured clinician review of the final model's projected prototypes, finding that they are rated as representative and clear. ProtoECGNet shows that prototype learning can be effectively scaled to complex, multi-label time-series classification, offering a practical path toward transparent and trustworthy deep learning models for clinical decision support.","PeriodicalId":74504,"journal":{"name":"Proceedings of machine learning research","volume":"298 ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2025-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12700622/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145758662","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Does Domain-Specific Retrieval Augmented Generation Help LLMs Answer Consumer Health Questions? 特定领域检索增强生成有助于法学硕士回答消费者健康问题吗？

Proceedings of machine learning research Pub Date : 2025-08-01

Chase M Fensore, Rodrigo M Carrillo-Larco, Megha K Shah, Joyce C Ho

{"title":"Does Domain-Specific Retrieval Augmented Generation Help LLMs Answer Consumer Health Questions?","authors":"Chase M Fensore, Rodrigo M Carrillo-Larco, Megha K Shah, Joyce C Ho","doi":"","DOIUrl":"","url":null,"abstract":"While large language models (LLMs) have shown impressive performance on medical benchmarks, there remains uncertainty about whether retrieval-augmented generation (RAG) meaningfully improves their ability to answer consumer health questions. In this study, we systematically evaluate vanilla LLMs against RAG-enhanced approaches using the NIDDK portion of the MedQuAD dataset. We compare four open-source LLMs in both vanilla and RAG configurations, assessing performance through automated metrics, LLM-based evaluation, and clinical validation. Surprisingly, we find that vanilla LLM approaches consistently outperform RAG variants across both quantitative metrics (BLEU, ROUGE, BERTScore) and qualitative assessments. The relatively low retrieval performance (Precision@5 = 0.15) highlights fundamental challenges in implementing effective RAG systems for medical question-answering, even with carefully curated questions. While RAG showed competitive performance in specific areas like scientific consensus and harm reduction, our findings suggest that successful implementation of RAG for consumer health question-answering requires more sophisticated approaches than simple retrieval and prompt engineering. These results contribute to the ongoing discussion about the role of retrieval augmentation in medical AI systems and highlight the need for medical-specific RAG infrastructure to enhance medical question-answering systems.","PeriodicalId":74504,"journal":{"name":"Proceedings of machine learning research","volume":"298 ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2025-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC13250811/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"148229676","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Balancing Interpretability and Flexibility in Modeling Diagnostic Trajectories with an Embedded Neural Hawkes Process Model. 利用嵌入式神经霍克斯过程模型平衡诊断轨迹建模的可解释性和灵活性。

Proceedings of machine learning research Pub Date : 2025-08-01

Yuankang Zhao, Matthew M Engelhard

{"title":"Balancing Interpretability and Flexibility in Modeling Diagnostic Trajectories with an Embedded Neural Hawkes Process Model.","authors":"Yuankang Zhao, Matthew M Engelhard","doi":"","DOIUrl":"","url":null,"abstract":"The Hawkes process (HP) is commonly used to model event sequences with self-reinforcing dynamics, including electronic health records (EHRs). Traditional HPs capture self-reinforcement via parametric impact functions that can be inspected to understand how each event modulates the intensity of others. Neural network-based HPs offer greater flexibility, resulting in improved fit and prediction performance, but at the cost of interpretability, which is often critical in healthcare. In this work, we aim to understand and improve upon this tradeoff. We propose a novel HP formulation in which impact functions are modeled by defining a flexible impact kernel, instantiated as a neural network, in event embedding space, which allows us to model large-scale event sequences with many event types. This approach is more flexible than traditional HPs yet more interpretable than other neural network approaches, and allows us to explicitly trade flexibility for interpretability by adding transformer encoder layers to further contextualize the event embeddings. Results show that our method accurately recovers impact functions in simulations, achieves competitive performance on MIMIC-IV procedure dataset, and gains clinically meaningful interpretation on Duke-EHR with children diagnosis dataset even without transformer layers. This suggests that our flexible impact kernel is often sufficient to capture self-reinforcing dynamics in EHRs and other data effectively, implying that interpretability can be maintained without loss of performance.","PeriodicalId":74504,"journal":{"name":"Proceedings of machine learning research","volume":"298 ","pages":""},"PeriodicalIF":0.0,"publicationDate":"2025-08-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC12646569/pdf/","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145643662","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Iterative Learning of Computable Phenotypes for Treatment Resistant Hypertension using Large Language Models. 使用大语言模型对治疗顽固性高血压的可计算表型进行迭代学习。

Proceedings of machine learning research Pub Date : 2025-08-01

Guilherme Seidyo Imai Aldeia, Daniel S Herman, William G La Cava

引用次数: 0

Stage-Aware Event-Based Modeling (SA-EBM) for Disease Progression. 基于阶段感知事件的疾病进展模型（SA-EBM）

Proceedings of machine learning research Pub Date : 2025-08-01

Hongtao Hao, Vivek Prabhakaran, Veena A Nair, Nagesh Adluru, Joseph L Austerweil

引用次数: 0