Skip to main content
. 2021 Apr 20;95(1132):20201107. doi: 10.1259/bjr.20201107

Figure 4.

Figure 4.

Overview of four key categories of evaluation metrics (overlap, distance, error and similarity) used to evaluate the performance of deep learning methods in medical image analysis. Each category contains brief descriptions and mathematical formulations for some common metrics. In these equations, ‘x’ and ‘y’ denote the prediction and target of any deep learning task, respectively.