Temporal Phenotyping using Deep Predictive Clustering of Disease Progression

Changhee Lee, Mihaela Van Der Schaar
Proceedings of the 37th International Conference on Machine Learning, PMLR 119:5767-5777, 2020.

Abstract

Due to the wider availability of modern electronic health records, patient care data is often being stored in the form of time-series. Clustering such time-series data is crucial for patient phenotyping, anticipating patients’ prognoses by identifying “similar” patients, and designing treatment guidelines that are tailored to homogeneous patient subgroups. In this paper, we develop a deep learning approach for clustering time-series data, where each cluster comprises patients who share similar future outcomes of interest (e.g., adverse events, the onset of comorbidities). To encourage each cluster to have homogeneous future outcomes, the clustering is carried out by learning discrete representations that best describe the future outcome distribution based on novel loss functions. Experiments on two real-world datasets show that our model achieves superior clustering performance over state-of-the-art benchmarks and identifies meaningful clusters that can be translated into actionable information for clinical decision-making.

Cite this Paper

BibTeX
@InProceedings{pmlr-v119-lee20h, title = {Temporal Phenotyping using Deep Predictive Clustering of Disease Progression}, author = {Lee, Changhee and Van Der Schaar, Mihaela}, booktitle = {Proceedings of the 37th International Conference on Machine Learning}, pages = {5767--5777}, year = {2020}, editor = {III, Hal Daumé and Singh, Aarti}, volume = {119}, series = {Proceedings of Machine Learning Research}, month = {13--18 Jul}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v119/lee20h/lee20h.pdf}, url = {https://proceedings.mlr.press/v119/lee20h.html}, abstract = {Due to the wider availability of modern electronic health records, patient care data is often being stored in the form of time-series. Clustering such time-series data is crucial for patient phenotyping, anticipating patients’ prognoses by identifying “similar” patients, and designing treatment guidelines that are tailored to homogeneous patient subgroups. In this paper, we develop a deep learning approach for clustering time-series data, where each cluster comprises patients who share similar future outcomes of interest (e.g., adverse events, the onset of comorbidities). To encourage each cluster to have homogeneous future outcomes, the clustering is carried out by learning discrete representations that best describe the future outcome distribution based on novel loss functions. Experiments on two real-world datasets show that our model achieves superior clustering performance over state-of-the-art benchmarks and identifies meaningful clusters that can be translated into actionable information for clinical decision-making.} }
Endnote
%0 Conference Paper %T Temporal Phenotyping using Deep Predictive Clustering of Disease Progression %A Changhee Lee %A Mihaela Van Der Schaar %B Proceedings of the 37th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2020 %E Hal Daumé III %E Aarti Singh %F pmlr-v119-lee20h %I PMLR %P 5767--5777 %U https://proceedings.mlr.press/v119/lee20h.html %V 119 %X Due to the wider availability of modern electronic health records, patient care data is often being stored in the form of time-series. Clustering such time-series data is crucial for patient phenotyping, anticipating patients’ prognoses by identifying “similar” patients, and designing treatment guidelines that are tailored to homogeneous patient subgroups. In this paper, we develop a deep learning approach for clustering time-series data, where each cluster comprises patients who share similar future outcomes of interest (e.g., adverse events, the onset of comorbidities). To encourage each cluster to have homogeneous future outcomes, the clustering is carried out by learning discrete representations that best describe the future outcome distribution based on novel loss functions. Experiments on two real-world datasets show that our model achieves superior clustering performance over state-of-the-art benchmarks and identifies meaningful clusters that can be translated into actionable information for clinical decision-making.
APA
Lee, C. & Van Der Schaar, M.. (2020). Temporal Phenotyping using Deep Predictive Clustering of Disease Progression. Proceedings of the 37th International Conference on Machine Learning, in Proceedings of Machine Learning Research 119:5767-5777 Available from https://proceedings.mlr.press/v119/lee20h.html.

Related Material