Online Unsupervised Representation Learning of Waveforms in the Intensive Care Unit via a Novel Cooperative Framework: Spatially Resolved Temporal Networks (SpaRTEn)
Abstract
Univariate high-frequency time series are dominant data sources for many medical, economic and environmental applications. In many of these domains, the time series are tied to real-time changes in state. In the intensive care unit, for example, changes and intracranial pressure waveforms can indicate whether a patient is developing decreased blood perfusion to the brain during a stroke, for example. However, most representation learning to resolve states is conducted in an offline, batch-dependent manner. In high frequency time-series, high intra-state and inter-sample variability makes offline, batch-dependent learning a relatively difficult task. Hence, we propose Spatial Resolved Temporal Networks (SpaRTeN), a novel composite deep learning model for online, unsupervised representation learning through a spatially constrained latent space. SpaRTeN maps waveforms to states, and learns time-dependent representations of each state. Our key contribution is that we generate clinically relevant representations of each state for intracranial pressure waveforms.
Cite
Text
Gulamali et al. "Online Unsupervised Representation Learning of Waveforms in the Intensive Care Unit via a Novel Cooperative Framework: Spatially Resolved Temporal Networks (SpaRTEn)." Proceedings of the 8th Machine Learning for Healthcare Conference, 2023.Markdown
[Gulamali et al. "Online Unsupervised Representation Learning of Waveforms in the Intensive Care Unit via a Novel Cooperative Framework: Spatially Resolved Temporal Networks (SpaRTEn)." Proceedings of the 8th Machine Learning for Healthcare Conference, 2023.](https://mlanthology.org/mlhc/2023/gulamali2023mlhc-online/)BibTeX
@inproceedings{gulamali2023mlhc-online,
title = {{Online Unsupervised Representation Learning of Waveforms in the Intensive Care Unit via a Novel Cooperative Framework: Spatially Resolved Temporal Networks (SpaRTEn)}},
author = {Gulamali, Faris and Sawant, Ashwin and Hofer, Ira and Levin, Matthew and Charney, Alexander and Singh, Karandeep and Glicksberg, Benjamin and Nadkarni, Girish},
booktitle = {Proceedings of the 8th Machine Learning for Healthcare Conference},
year = {2023},
pages = {230-247},
volume = {219},
url = {https://mlanthology.org/mlhc/2023/gulamali2023mlhc-online/}
}