Sex-specific computationally generated biomarkers for cardiovascular risk stratification post acute coronary syndrome

• Main Author: Chong Rodriguez, Alicia
• Other Authors: Collin M. Stultz.
• Format: Others
• Language: English
• Published: Massachusetts Institute of Technology 2018
• Subjects: Engineering and Management Program.; Integrated Design and Management Program.; Electrical Engineering and Computer Science.
• Online Access: http://hdl.handle.net/1721.1/118555

Thesis: S.M. in Engineering and Management, Massachusetts Institute of Technology, System Design and Management Program, 2018. === Thesis: S.M., Massachusetts Institute of Technology, Department of Electrical Engineering and Computer Science, 2018. === Cataloged from PDF version of thesis. === Includes bibliographical references (pages 46-48). === Women who present with symptoms consistent with an Acute Coronary Syndrome (ACS) are often under-diagnosed and under-represented in clinical trials. Moreover, there are data to suggest that women with cardiovascular disease have worse outcomes, poorer prognoses, and higher mortality rates than men. Determining the risk of future adverse cardiovascular events for women who have previously suffered an ACS is therefore a problem of paramount importance in the field of cardiovascular medicine. The identification of high-risk patient subgroups typically begins with an evaluation of the patient's history, physical exam, and the surface electrocardiogram (ECG). Indeed, the ECG plays a central role in the assessment and management of patients post ACS. In this study, we develop and test a technique for automatically assessing the risk of death in women who presented with an ACS. The method combines both patient history and an automated analysis of the surface ECG to accurately quantify that patient's future risk. The clustering of patients into subgroups, each having a different level of risk, is used to develop an algorithm to quantify the risk of new patients who present with an ACS. In this work, a comprehensive comparison between clustering female only data and traditional, female and male data is demonstrated as risk stratification methodologies for learning the significance or impact of our test and its inputs. The model trained on the entire population always performs worse for female population and the model trained only on female patients always provides a better performance for these patients. Comparing to existing risk scores, the female-specific model performs better. === by Alicia Chong Rodriguez. === S.M. in Engineering and Management === S.M.