TY - GEN
T1 - Outcomes-Driven Clinical Phenotyping in Cardiogenic Shock using a Mixture of Experts
AU - Hurley, Nathan C.
AU - Berkowitz, Alyssa
AU - Masoudi, Frederick
AU - Ross, Joseph
AU - Desai, Nihar
AU - Shah, Nilay
AU - Dhruva, Sanket
AU - Mortazavi, Bobak J.
N1 - Funding Information:
Funding for this research was provided by a Center of Excellence in Regulatory Science and Innovation (CERSI) grant to Yale University and Mayo Clinic from the FDA (U01FD005938). This research was supported by the American College of Cardiology Foundation’s National Cardiovascular Data Registry (NCDR). The views expressed in this presentation represent those of the authors, and do not necessarily represent the official views of the NCDR or its associated professional societies identified at CVQuality.ACC.org. For more information go to CVQuality.ACC.org or email ncdrresearch@acc.org
Funding Information:
This research was provided by a Center of Excellence in Regulatory Science and Innovation (CERSI) grant to Yale University and Mayo Clinic from the FDA (U01FD005938). This research was supported by the American College of Cardiology Foundation's National Cardiovascular Data Registry (NCDR). The views expressed in this presentation represent those of the authors, and do not necessarily represent the official views of the NCDR or its associated professional societies identified at CVQuality.ACC.org. For more information go to CVQuality.ACC.org or email ncdrresearch@acc.org
Publisher Copyright:
© 2021 IEEE
PY - 2021
Y1 - 2021
N2 - Cardiogenic shock (CS) is a deadly and complicated illness. Despite extensive research into its treatment, mortality remains high and has not decreased over time. Patients suffering from CS are highly heterogeneous. Developing an understanding of phenotypes among these patients is crucial for understanding this disease and appropriate treatments for individual patients. In this work, we develop a deep mixture of experts approach to jointly find phenotypes among patients with CS while simultaneously estimating their risk of in-hospital mortality. This model is applied to a cohort of 28,304 patients with CS, predicting in-hospital mortality with an AUROC of 0.85 ± 0.01 and discovering five phenotypes among the population. This approach allows for grouping patients in clinical clusters with different rates of device utilization and different risk of mortality. This approach jointly finds phenotypes within a clinical population and in modeling risk among that population.
AB - Cardiogenic shock (CS) is a deadly and complicated illness. Despite extensive research into its treatment, mortality remains high and has not decreased over time. Patients suffering from CS are highly heterogeneous. Developing an understanding of phenotypes among these patients is crucial for understanding this disease and appropriate treatments for individual patients. In this work, we develop a deep mixture of experts approach to jointly find phenotypes among patients with CS while simultaneously estimating their risk of in-hospital mortality. This model is applied to a cohort of 28,304 patients with CS, predicting in-hospital mortality with an AUROC of 0.85 ± 0.01 and discovering five phenotypes among the population. This approach allows for grouping patients in clinical clusters with different rates of device utilization and different risk of mortality. This approach jointly finds phenotypes within a clinical population and in modeling risk among that population.
KW - Cardiology
KW - Knowledge engineering
KW - Machine learning
KW - Medical information systems
KW - Neural networks
UR - http://www.scopus.com/inward/record.url?scp=85125497539&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85125497539&partnerID=8YFLogxK
U2 - 10.1109/BHI50953.2021.9508568
DO - 10.1109/BHI50953.2021.9508568
M3 - Conference contribution
AN - SCOPUS:85125497539
T3 - BHI 2021 - 2021 IEEE EMBS International Conference on Biomedical and Health Informatics, Proceedings
BT - BHI 2021 - 2021 IEEE EMBS International Conference on Biomedical and Health Informatics, Proceedings
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 2021 IEEE EMBS International Conference on Biomedical and Health Informatics, BHI 2021
Y2 - 27 July 2021 through 30 July 2021
ER -