Use of Neural Networks to Predict Adverse Outcomes from Acute Coronary Syndrome for Male and Female Patients

Neural networks have been used to examine a set of thirteen objective features and a single subjective physician's assessment for emergency room patients with symptoms possibly indicative of acute coronary syndrome (ACS). The objective data is information routinely collected during triage. The neural networks were used to fuse the disparate types of information with the goal of forecasting thirty-day adverse patient outcome. Results were evaluated using receiver operating characteristic curves describing the outcomes of the nets, both using only objective features and including the subjective physician's assessment. These results, based on all patient data, are compared to those obtained using neural networks trained on information from male and female patients separately. While preliminary, the results of this continuing study are significant from the perspective of potential use of the intelligent fusion of biomedical informatics to aid the physician in prescribing treatment necessary to prevent serious adverse outcome from ACS.