Learning from undercoded clinical records for automated International Classification of Diseases (ICD) coding
Journal of the American Medical Informatics Association · 2023
DOI: 10.1093/jamia/ocac230 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of the American Medical Informatics Association, 2023. Verfasst von Yucheng Jin und 9 weiteren Autor:innen. Beteiligte Einrichtung: Shanghai Key Laboratory of Data Science, School of Computer Science, Fudan University, Shanghai, China. Hauptthemen laut MEDLINE: International Classification of Diseases, Electronic Health Records. Die Arbeit wurde bisher 8-mal zitiert (Crossref) und nennt 28 Literaturangaben. Der Volltext ist frei in PubMed Central verfügbar. Gefördert von: China Shanghai Science and Technology Development Fund, Major Key Project of Peng Cheng Laboratory, China Shanghai Municipal Health Commission Advanced Appropriate Technology.
Abstract (Originalsprache)
Objectives
To develop an unbiased objective for learning automatic coding algorithms from clinical records annotated with only partial relevant International Classification of Diseases codes, as annotation noise in undercoded clinical records used as training data can mislead the learning process of deep neural networks. Materials and Methods We use Medical Information Mart for Intensive Care III as our dataset. We employ positive-unlabeled learning to achieve unbiased loss estimation, which is free of misleading training signal. We then utilize reweighting mechanism to compensate for the imbalance between positive and negative samples. To further close the performance gap caused by poor quality annotation, we integrate the supervision provided by the automatic annotation tool Medical Concept Annotation Toolkit which can ease the heavy burden of manual validation.
Results
Our benchmarking results show that positive-unlabeled learning with reweighting outperforms competitive baseline methods over a range of missing label ratios. Integrating supervision provided by annotation tool further boosted the performance.
Discussion
Considering the annotation noise and severe imbalance, unbiased loss estimation and reweighting mechanism are both important for learning from undercoded clinical records. Unbiased loss requires the estimation of false negative ratios and estimation through trained models is practical and competitive.
Conclusions
The combination of positive-unlabeled learning with reweighting and supervision provided by the annotation tool is a promising solution to learn from undercoded clinical records.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Shanghai Key Laboratory of Data Science, School of Computer Science, Fudan University, Shanghai, China.
- Peng Cheng Laboratory, Shenzhen, China.
- Department of Geriatrics, Yueyang Hospital of Integrated Traditional Chinese Medicine and Western Medicine, Shanghai University of Traditional Chinese Medicine, Shanghai, China.
- Medical Device Regulatory Research Center/Precision Medicine Research Center, West China Hospital of Sichuan University, Sichuan, China.
- Department of Epidemiology, Harvard T.H. Chan School of Public Health, Boston, Massachusetts, USA.
- Department of Computer Science and Engineering, Lehigh University, Bethlehem, Pennsylvania, USA.
ORCID-Profile: Yucheng Jin, Yun Xiong, Dan Shi, Yifei Lin, Lifang He, Joseph M Plasek, Li Zhou, David W Bates, Chunlei Tang
Förderung
- China Shanghai Science and Technology Development Fund
- Major Key Project of Peng Cheng Laboratory
- China Shanghai Municipal Health Commission Advanced Appropriate Technology
Zitieren
APA
Jin, Y., Xiong, Y., Shi, D., Lin, Y., He, L., Zhang, Y., Plasek, J. M., Zhou, L., Bates, D. W., & Tang, C. (2023). Learning from undercoded clinical records for automated International Classification of Diseases (ICD) coding. Journal of the American Medical Informatics Association. https://doi.org/10.1093/jamia/ocac230
Vancouver
Jin Y, Xiong Y, Shi D, Lin Y, He L, Zhang Y, et al. Learning from undercoded clinical records for automated International Classification of Diseases (ICD) coding. Journal of the American Medical Informatics Association. 2023. doi:10.1093/jamia/ocac230
BibTeX
@article{Jin2023,
author = {Jin, Yucheng and Xiong, Yun and Shi, Dan and Lin, Yifei and He, Lifang and Zhang, Yao and Plasek, Joseph M and Zhou, Li and Bates, David W and Tang, Chunlei},
title = {Learning from undercoded clinical records for automated International Classification of Diseases (ICD) coding},
journal = {Journal of the American Medical Informatics Association},
year = {2023},
doi = {10.1093/jamia/ocac230}
}
EUVIMED