Fig. 2: DeepNeo accurately segments lumen, stent, and neointima. | Communications Medicine

Fig. 2: DeepNeo accurately segments lumen, stent, and neointima.

From: Deep learning model DeepNeo predicts neointimal tissue characterization using optical coherence tomography

Fig. 2

Beeswarm plot with boxplot, with median values (central horizontal black line), boxes extend from the 25th to the 75th percentile of scores generated by 5-fold cross-validation on n = 305 images from 42 patients. Whiskers represent the range from the minimum to the maximum values within 1.5 times the interquartile range from the 25th and 75th percentiles (a). Good, average, and low-performing samples with respect to the average Dice score (dsc) of an image (b). The Dice score is calculated as the area of overlap between labeled ground truth and prediction, ranging from 0 to 1 (0 indicating no overlap and 1 complete overlap between prediction and ground truth).

Back to article page