Artificial Intelligence

Representative examples of the attention heatmaps generated using Grad-CAM method for (a) COVID-19, (b) CAP, and (c) Non-Pneumonia. The heatmaps are standard Jet colormap and overlapped on the original image, the red color highlights the activation region associated with the predicted class. COVID-19 = coronavirus disease 2019, CAP = community acquired pneumonia. Image courtesy of the journal Radiology

Figure 1: Depiction of the fully automated CT biomarkers tools used in this study. (A) Schematic depiction of the automated process for assessing fat, muscle, liver, aortic calcification, and bone from original abdominal CT scan data. (B) Case example in an asymptomatic 52-year-old man undergoing CT for colorectal cancer screening. At the time of CT screening, he had a body-mass index of 27·3 and Framingham risk score of 5% (low risk). However, several CT-based metabolic markers were indicative of underlying disease. Multivariate Cox model prediction based on these three CT-based results put the risk of cardiovascular event at 19% within 2 years, at 40% within 5 years, and at 67% within 10 years, and the risk of death at 4% within 2 years, 11% within 5 years, and 27% within 10 years. At longitudinal clinical follow-up, the patient suffered an acute myocardial infarction 3 years after this initial CT and died 12 years after CT at the age of 64 years. (C) Contrast-enhanced CT performed 7 months before death for minor trauma was interpreted as negative but does show significant progression of vascular calcification, visceral fat, and hepatic steatosis. HU=Hounsfield units.

TeraRecon's End-to-End AI Ecosystem