HiCu: Leveraging Hierarchy for Curriculum Learning in Automated ICD Coding

Weiming Ren, Ruijing Zeng, Tongzi Wu, Tianshu Zhu, Rahul G. Krishnan
Proceedings of the 7th Machine Learning for Healthcare Conference, PMLR 182:198-223, 2022.

Abstract

There are several opportunities for automation in healthcare that can improve clinician throughput. One such example is assistive tools to document diagnosis codes when clinicians write notes. We study the automation of medical code prediction using curriculum learning, which is a training strategy for machine learning models that gradually increases the hardness of the learning tasks from easy to difficult. One of the challenges in curriculum learning is the design of curricula – i.e., in the sequential design of tasks that gradually increase in difficulty. We propose Hierarchical Curriculum Learning (HiCu), an algorithm that uses graph structure in the space of outputs to design curricula for multi-label classification. We create curricula for multi-label classification models that predict ICD diagnosis and procedure codes from natural language descriptions of patients. By leveraging the hierarchy of ICD codes, which groups diagnosis codes based on various organ systems in the human body, we find that our proposed curricula improve the generalization of neural network-based predictive models across recurrent, convolutional, and transformer-based architectures. Our code is available at https://github.com/wren93/HiCu-ICD.

Cite this Paper

BibTeX
@InProceedings{pmlr-v182-ren22a, title = {HiCu: Leveraging Hierarchy for Curriculum Learning in Automated ICD Coding}, author = {Ren, Weiming and Zeng, Ruijing and Wu, Tongzi and Zhu, Tianshu and Krishnan, Rahul G.}, booktitle = {Proceedings of the 7th Machine Learning for Healthcare Conference}, pages = {198--223}, year = {2022}, editor = {Lipton, Zachary and Ranganath, Rajesh and Sendak, Mark and Sjoding, Michael and Yeung, Serena}, volume = {182}, series = {Proceedings of Machine Learning Research}, month = {05--06 Aug}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v182/ren22a/ren22a.pdf}, url = {https://proceedings.mlr.press/v182/ren22a.html}, abstract = {There are several opportunities for automation in healthcare that can improve clinician throughput. One such example is assistive tools to document diagnosis codes when clinicians write notes. We study the automation of medical code prediction using curriculum learning, which is a training strategy for machine learning models that gradually increases the hardness of the learning tasks from easy to difficult. One of the challenges in curriculum learning is the design of curricula – i.e., in the sequential design of tasks that gradually increase in difficulty. We propose Hierarchical Curriculum Learning (HiCu), an algorithm that uses graph structure in the space of outputs to design curricula for multi-label classification. We create curricula for multi-label classification models that predict ICD diagnosis and procedure codes from natural language descriptions of patients. By leveraging the hierarchy of ICD codes, which groups diagnosis codes based on various organ systems in the human body, we find that our proposed curricula improve the generalization of neural network-based predictive models across recurrent, convolutional, and transformer-based architectures. Our code is available at https://github.com/wren93/HiCu-ICD.} }
Endnote
%0 Conference Paper %T HiCu: Leveraging Hierarchy for Curriculum Learning in Automated ICD Coding %A Weiming Ren %A Ruijing Zeng %A Tongzi Wu %A Tianshu Zhu %A Rahul G. Krishnan %B Proceedings of the 7th Machine Learning for Healthcare Conference %C Proceedings of Machine Learning Research %D 2022 %E Zachary Lipton %E Rajesh Ranganath %E Mark Sendak %E Michael Sjoding %E Serena Yeung %F pmlr-v182-ren22a %I PMLR %P 198--223 %U https://proceedings.mlr.press/v182/ren22a.html %V 182 %X There are several opportunities for automation in healthcare that can improve clinician throughput. One such example is assistive tools to document diagnosis codes when clinicians write notes. We study the automation of medical code prediction using curriculum learning, which is a training strategy for machine learning models that gradually increases the hardness of the learning tasks from easy to difficult. One of the challenges in curriculum learning is the design of curricula – i.e., in the sequential design of tasks that gradually increase in difficulty. We propose Hierarchical Curriculum Learning (HiCu), an algorithm that uses graph structure in the space of outputs to design curricula for multi-label classification. We create curricula for multi-label classification models that predict ICD diagnosis and procedure codes from natural language descriptions of patients. By leveraging the hierarchy of ICD codes, which groups diagnosis codes based on various organ systems in the human body, we find that our proposed curricula improve the generalization of neural network-based predictive models across recurrent, convolutional, and transformer-based architectures. Our code is available at https://github.com/wren93/HiCu-ICD.
APA
Ren, W., Zeng, R., Wu, T., Zhu, T. & Krishnan, R.G.. (2022). HiCu: Leveraging Hierarchy for Curriculum Learning in Automated ICD Coding. Proceedings of the 7th Machine Learning for Healthcare Conference, in Proceedings of Machine Learning Research 182:198-223 Available from https://proceedings.mlr.press/v182/ren22a.html.

Related Material