Fig. 9 | Scientific Reports

Fig. 9

From: Integrating radiomic texture analysis and deep learning for automated myocardial infarction detection in cine-MRI

Fig. 9

Feature space visualization using Principal Component Analysis (PCA, left) and t-distributed Stochastic Neighbor Embedding (t-SNE, right). Each point represents a subject, color-coded by class (blue: Normal, red: MI). PCA shows partial separation, while t-SNE reveals more apparent infarcted versus healthy sample clustering, indicating effective radiomic feature discrimination within the hybrid U-Net framework.

Back to article page