CLIR-Bench: Benchmarking Multimodal Question Answering over Irregular Clinical Time Series

2026-07-10

AI Summary

Existing clinical question-answering benchmarks struggle to evaluate models’ ability to locate temporal evidence and reason over sparse, asynchronous, and irregularly sampled clinical time-series data. To address this gap, this work presents the first systematic multimodal question-answering benchmark tailored for irregular clinical time series, comprising 6,600 questions derived from de-identified ICU records through a four-stage generation pipeline. The benchmark spans 11 clinical variables and four core reasoning dimensions, with each question explicitly linked to temporal evidence and task-specific answer derivation rules, enabling fine-grained assessment of models’ evidence utilization capabilities. Experimental results demonstrate that current general-purpose models fail to effectively retrieve and reason over such sparse temporal evidence, underscoring the need for specialized methods for irregular time-series reasoning. The dataset and code are publicly released.

Abstract

Clinical time series are central to patient monitoring, risk assessment, and clinical decision support. However, they are often sparse, irregularly sampled, and asynchronous, making it difficult for models to identify the temporal evidence required for clinical Question Answering (QA). Existing benchmarks primarily focus on regularly sampled time-series QA or medical QA over static data, and therefore rarely assess whether models can faithfully ground their answers in irregular temporal observations. To fill this gap, we introduce CLIR-Bench, a benchmark for irregular clinical time series QA constructed from de-identified ICU records through a principled four-stage pipeline. CLIR-Bench contains 6,600 QA instances spanning 11 clinical variables, organized into four capability dimensions and 11 tasks. Each question is linked to explicit temporal evidence and task-specific answer derivation rules, enabling evaluation of both answer accuracy and evidence use. Experiments show that existing generalist models struggle to retrieve and reason over sparse clinical evidence, highlighting the need for stronger irregular time-series reasoning methods. Our code and data are available at https://huggingface.co/datasets/winall/CLIR-Bench.

Problem

• clinical time series
• irregular sampling
• multimodal question answering
• temporal evidence
• clinical QA

Innovation

• irregular time series
• clinical question answering
• temporal reasoning
• multimodal benchmark
• evidence grounding