Fig. 2: Performance of the screening and diagnostic models in internal and external testing. | Nature Medicine

Fig. 2: Performance of the screening and diagnostic models in internal and external testing.

From: Screening and diagnosis of cardiovascular disease using artificial intelligence-enabled cardiac magnetic resonance imaging

Fig. 2

a, ROCs for the screening of cardiac anomalies for the primary internal test dataset (blue, n = 7,900) and external test dataset (red, n = 1,819). The screening model is derived from 4CH cine and SAX cine. b, The diagnostic performance for the internal test dataset (yellow, n = 6,650) and external test dataset (blue, n = 1,416). The diagnostic model takes cine (4CH and SAX) and LGE as combined inputs. c, A confusion matrix for the predictions of the AI diagnostic model versus the ground truth over the entire CVD cohort (n = 8,066). The percentage of all possible predictions in each CVD class is displayed on a color gradient scale. d, ROCs for the diagnosis of CVD classes for the internal set and external set.

Back to article page