Framework for precision phenotyping of heart failure patients using clinical data at rest and exercise

Autor: Brian Carlson, Andrew Meyer, Filip Ježek, Edith Jones, Xinwei Hua, Scott Hummel, Yida Tang, Daniel Beard
Rok vydání: 2023
Předmět:
Zdroj: Physiology. 38
ISSN: 1548-9221
1548-9213
DOI: 10.1152/physiol.2023.38.s1.5796077
Popis: Precision phenotyping of mechanistic differences between and within different types of heart failure is a valuable tool not only to understand the range of underlying cardiovascular dysfunction but to identify patients who could benefit from a targeted treatment or procedure. Our group previously has used standard clinical hemodynamic measures at rest coupled with a computational model of the cardiovascular system to identify differences between patients diagnosed with heart failure with preserved ejection fraction (HFpEF). Even though this effort has proven fruitful, many patients with HFpEF exhibit symptoms only during exercise. We hypothesize that incorporating exercise metrics into our approach will improve discrimination between HFpEF phenotypes. In this study we have expanded the pool of clinical data to include cardiopulmonary exercise testing (CPET) data and added to the complexity of our cardiovascular system model to simulate exercise. Data were collected prospectively from both Peking University Third Hospital and the University of Michigan and include measures from right heart catheterization (RHC) at rest and at exercise along with transthoracic echocardiography (TTE) at rest. The measures for each patient are used to identify our expanded cardiovascular systems model which now can inform parameters related to the strength of sarcomeric contraction in the left, septal and right ventricular walls, left and right atrial elastance in addition to predicting left ventricular and right ventricular cardiac power at rest and exercise. Based on the preliminary data acquired, we present a framework that can quantify underlying differences between HFpEF patients using unsupervised machine learning to find distinct HFpEF phenogroups, using clinical data both at rest and during exercise. Michigan Medicine/Peking University Health Science Center Joint Institute for Translational and Clinical Research This is the full abstract presented at the American Physiology Summit 2023 meeting and is only available in HTML format. There are no additional versions or additional content available for this abstract. Physiology was not involved in the peer review process.
Databáze: OpenAIRE
Externí odkaz: