[공지사항]을 빙자한 안부와 근황 
Show more

Daily Arxiv

This is a page that curates AI-related papers published worldwide.
All content here is summarized using Google Gemini and operated on a non-profit basis.
Copyright for each paper belongs to the authors and their institutions; please make sure to credit the source when sharing.

DENSE: Longitudinal Progress Note Generation with Temporal Modeling of Heterogeneous Clinical Notes Across Hospital Visits

Created by
  • Haebom

Author

Garapati Keerthana, Manik Gupta

Outline

DENSE is a system designed to solve the problem of lack of progress records in electronic health records (EHR). In EHR datasets lacking progress records, such as the MIMIC-III dataset, DENSE chronologically sorts various types of notes, retrieves clinically relevant information, and generates progress records using a large-scale language model (LLM). It is designed to mimic the way physicians refer to past medical records, and aims to generate progress records with high temporal consistency to improve subsequent tasks such as summarization, predictive modeling, and clinical decision support. The evaluation results show that the generated progress records have higher temporal consistency (temporal alignment ratio 1.089) than the original records.

Takeaways, Limitations

•
Takeaways:
◦
Presenting a practical approach to address the lack of progress records in EHR datasets
◦
Presenting the possibility of improving the efficiency and accuracy of clinical record creation using LLM
◦
Contributes to improving various follow-up tasks such as summary, predictive modeling, and clinical decision support
◦
Improve longitudinal patient information analysis by creating a temporally consistent progress record
•
Limitations:
◦
The evaluation dataset is limited to patients with complete progress records, which may not fully reflect the diversity of real-world clinical settings.
◦
The performance of DENSE depends on the LLM and clinical information retrieval strategy used, and improvements in their performance directly affect the performance of DENSE.
◦
Further research is needed on applicability and scalability in real clinical settings.
◦
Consideration of ethical and legal issues (e.g., patient privacy, medical liability) is required.
👍
Made with Slashpage