PMOA-TTS: Introducing the PubMed Open Access Textual Times Series Corpus

2025-05-23

AI Summary

This study addresses the scarcity of large-scale, high-precision temporally annotated data in clinical narratives. We introduce the first open-source PubMed-based clinical temporal corpus, automatically extracted from 125,000 open-access case reports, comprising over 5.6 million timestamped clinical events and structured patient trajectories. Methodologically, we propose an LLM-driven, multi-stage temporal event extraction framework—built upon prompt engineering pipelines with Llama 3.3 and DeepSeek-R1—and integrate heuristic filtering with semantic matching. Our key contribution is a novel three-dimensional clinical credibility evaluation framework, combining cosine similarity, concordance index (c-index), and area under the longitudinal time–concurrence curve (AULTC). The framework achieves an event matching rate of 80% (cosine ≥ 0.1) and temporal consistency c-index > 0.90. In downstream survival prediction, it attains a c-index of 0.82 ± 0.01, demonstrating strong predictive power of its temporal representations.

Technology Category

• Data Mining & Knowledge Management: Mining of Spatial, Temporal or Spatio-Temporal Data
• Natural Language Processing: Information Extraction
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning

Application Category

• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Web Mining and Content Analysis: Web data provenance, reliability, and authenticity
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

Understanding temporal dynamics in clinical narratives is essential for modeling patient trajectories, yet large-scale temporally annotated resources remain limited. We present PMOA-TTS, the first openly available dataset of 124,699 PubMed Open Access (PMOA) case reports, each converted into structured (event, time) timelines via a scalable LLM-based pipeline. Our approach combines heuristic filtering with Llama 3.3 to identify single-patient case reports, followed by prompt-driven extraction using Llama 3.3 and DeepSeek R1, resulting in over 5.6 million timestamped clinical events. To assess timeline quality, we evaluate against a clinician-curated reference set using three metrics: (i) event-level matching (80% match at a cosine similarity threshold of 0.1), (ii) temporal concordance (c-index>0.90), and (iii) Area Under the Log-Time CDF (AULTC) for timestamp alignment. Corpus-level analysis shows wide diagnostic and demographic coverage. In a downstream survival prediction task, embeddings from extracted timelines achieve time-dependent concordance indices up to 0.82 $pm$ 0.01, demonstrating the predictive value of temporally structured narratives. PMOA-TTS provides a scalable foundation for timeline extraction, temporal reasoning, and longitudinal modeling in biomedical NLP. The dataset is available at: https://huggingface.co/datasets/snoroozi/pmoa-tts .

Problem

• Lack of large-scale temporally annotated clinical narrative datasets
• Need for scalable methods to extract structured timelines from case reports
• Limited resources for temporal reasoning in biomedical NLP applications

Innovation

• LLM-based pipeline extracts clinical timelines
• Heuristic filtering with Llama 3.3
• Prompt-driven extraction using multiple models