Fig. 2.

Study Design and Overview. (A) The internal validation patient cohort presenting with non-ST-elevation myocardial infarction (NSTEMI, n = 63) and ST-elevation myocardial infarction (STEMI, n = 92) (B) Ultrasomics features were extracted and TDAView was used to cluster patients into three phenogroups: Cluster A, Cluster B, and Cluster C. The identified phenogroups were used to develop class labels for the external validation group using a supervised classifier. (C) The generated probabilities from the supervised classifier were used to predict mortality and illustrate the incremental value of ultrasomics features over GRACE 2.0. The supervised classifier was applied to the external validation group to develop class labels, which were used to predict findings on cardiac magnetic resonance, including acute infarct size