A multi-task deep learning model for the classification of Age-related Macular Degeneration

Abstract

Age-related Macular Degeneration (AMD) is a leading cause of blindness. Although the Age-Related Eye Disease Study group previously developed a 9-step AMD severity scale for manual classification of AMD severity from color fundus images, manual grading of images is time-consuming and expensive. Built on our previous work DeepSeeNet, we developed a novel deep learning model for automated classification of images into the 9-step scale. Instead of predicting the 9-step score directly, our approach simulates the reading center grading process. It first detects four AMD characteristics (drusen area, geographic atrophy, increased pigment, and depigmentation), then combines these to derive the overall 9-step score. Importantly, we applied multi-task learning techniques, which allowed us to train classification of the four characteristics in parallel, share representation, and prevent overfitting. Evaluation on two image datasets showed that the accuracy of the model exceeded the current state-of-the-art model by > 10%. Availability: https://github.com/ncbi-nlp/DeepSeeNet.

Figure 1:

The multi-task deep learning model for classification of age-related macular degeneration using the Age-Related Eye Disease Study Severity Scale.

Figure 2:

An example of (a) the original image, and (b) the pre-processed color fundus photograph.

Figure 3:

Confusion matrices for individual AMD characteristics.

Figure 4:

Performance of individual AMD characteristics.

Figure 5:

F1-score class difference (AREDS vs AREDS2). For example, the CNN model had 17.96% lower F1-score on class 4 compared AREDS to AREDS2.