Fig. 7: We generated visualizations using an improved version of gradient-weighted class activation mapping. | npj Digital Medicine

Fig. 7: We generated visualizations using an improved version of gradient-weighted class activation mapping.

From: Diagnosing pathologic myopia by identifying morphologic patterns using ultra widefield images with deep learning

Fig. 7

These visualizations show the qualitative predictions of RealMNet for presence of posterior staphyloma (NoPS or PS) and myopic maculopathy with five categories: no myopic retinal lesions (NoMRL), tessellated fundus only (TFO), diffuse chorioretinal atrophy (DCA), patchy chorioretinal atrophy (PCA), and macular atrophy (MA). By merging the heatmaps with the original images, we highlighted irregular attentive regions that correspond to diverse morphologic patterns found in different lesion categories when the model made decisions. These heatmaps provided a qualitative reference for clinicians when making further diagnoses.

Back to article page