A comparison of machine learning methods for risk stratification after acute coronary syndrome

Accurate risk stratification is essential for the proper management of patients after an acute coronary syndrome (ACS). Currently, the most widely accepted metrics for risk stratification are risk scores such as the Thrombolysis in Myocardial Infarction (TIMI) score and Global Registry of Acute Coronary Events (GRACE) score. However, prior work has shown that many patients who are not traditionally defined as high-risk by the TIMI or GRACE scores suffer adverse events such as cardiovascular death. We therefore wish to find a method of risk stratifying patients that has greater discriminatory ability than the existing scoring metrics. We wish to find a model that can assign a risk score using data that is routinely collected for patients during a hospital stay. Using a dataset of over 4200 patients, we developed logistic regression, neural network, and regression tree models to risk stratify patients for one-year cardiovascular death post ACS. The resulting models were highly predictive of risk compared to the TIMI score. Our findings highlight the efficacy of using machine learning models trained on commonly collected clinical data to risk stratify patients.

Department dc:contributor.department

Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science.

Grantor dc:publisher

Massachusetts Institute of Technology

Year dc:date.issued

2018

Author dc:creator

• Pavlick, Stephanie (Stephanie A.)

Advisor dc:contributor.advisor

• Collin M. Stultz.

• Electrical Engineering and Computer Science.

Statement dc:rights

• MIT theses are protected by copyright. They may be viewed, downloaded, or printed from this source but further reproduction or distribution in any format is prohibited without written permission.

Licence dc:rights.uri

• http://dspace.mit.edu/handle/1721.1/7582

Language dc:language.iso

eng

Handle dc:identifier.uri

http://hdl.handle.net/1721.1/119775

OAI identifier oai:identifier

oai:dspace.mit.edu:1721.1/119775