Multiloss Function Based Deep Convolutional Neural Network for Segmentation of Retinal Vasculature into Arterioles and Venules

Table 1

Performance measures applied in artery/vein classification.

Metric	Description
Mean Accuracy	the ratio of the correctly classified pixels averaged over the classes
Global accuracy	It measures the ratio of pixels accurately classified in the dataset.
Mean Intersection over Union (mIoU)	The similarity comparison between the class resulting from the prediction and the actual ground truth class
Weighted Intersection over Union (wIoU)	Performs IoU average weighted by the number of pixels in every class
Mean BF Score	This performance metric measures how accurate the boundaries of the predicted objects match the boundaries of the ground truth

TP, TN, FP, and FN are abbreviations of true positive, true negative, false positive, and false negative, respectively.