TY - GEN

T1 - Automated Medical Coding on MIMIC-III and MIMIC-IV

T2 - 46th International ACM SIGIR Conference on Research and Development in Information Retrieval, SIGIR 2023

AU - Edin, Joakim

AU - Junge, Alexander

AU - Havtorn, Jakob D.

AU - Borgholt, Lasse

AU - Maistro, Maria

AU - Ruotsalo, Tuukka

AU - Maaløe, Lars

N1 - Publisher Copyright: © 2023 Copyright held by the owner/author(s). Publication rights licensed to ACM.

PY - 2023

Y1 - 2023

N2 - Medical coding is the task of assigning medical codes to clinical free-text documentation. Healthcare professionals manually assign such codes to track patient diagnoses and treatments. Automated medical coding can considerably alleviate this administrative burden. In this paper, we reproduce, compare, and analyze state-of-the-art automated medical coding machine learning models. We show that several models underperform due to weak configurations, poorly sampled train-test splits, and insufficient evaluation. In previous work, the macro F1 score has been calculated sub-optimally, and our correction doubles it. We contribute a revised model comparison using stratified sampling and identical experimental setups, including hyperparameters and decision boundary tuning. We analyze prediction errors to validate and falsify assumptions of previous works. The analysis confirms that all models struggle with rare codes, while long documents only have a negligible impact. Finally, we present the first comprehensive results on the newly released MIMIC-IV dataset using the reproduced models. We release our code, model parameters, and new MIMIC-III and MIMIC-IV training and evaluation pipelines to accommodate fair future comparisons.

AB - Medical coding is the task of assigning medical codes to clinical free-text documentation. Healthcare professionals manually assign such codes to track patient diagnoses and treatments. Automated medical coding can considerably alleviate this administrative burden. In this paper, we reproduce, compare, and analyze state-of-the-art automated medical coding machine learning models. We show that several models underperform due to weak configurations, poorly sampled train-test splits, and insufficient evaluation. In previous work, the macro F1 score has been calculated sub-optimally, and our correction doubles it. We contribute a revised model comparison using stratified sampling and identical experimental setups, including hyperparameters and decision boundary tuning. We analyze prediction errors to validate and falsify assumptions of previous works. The analysis confirms that all models struggle with rare codes, while long documents only have a negligible impact. Finally, we present the first comprehensive results on the newly released MIMIC-IV dataset using the reproduced models. We release our code, model parameters, and new MIMIC-III and MIMIC-IV training and evaluation pipelines to accommodate fair future comparisons.

KW - Automated Medical Coding

KW - MIMIC

KW - Reproducibility

U2 - 10.1145/3539618.3591918

DO - 10.1145/3539618.3591918

M3 - Article in proceedings

AN - SCOPUS:85168694311

SP - 2572

EP - 2582

BT - SIGIR 2023 - Proceedings of the 46th International ACM SIGIR Conference on Research and Development in Information Retrieval

PB - Association for Computing Machinery, Inc.

Y2 - 23 July 2023 through 27 July 2023

ER -