A generalizable approach based on the <scp>U‐Net</scp> model for automatic intraretinal cyst segmentation in <scp>SD‐OCT</scp> images

In this article, we propose a new U-Net-based approach for intraretinal cyst segmentation across different vendors that improve some of the challenges faced by previous deep-based techniques. The proposed method has two main steps: (1) prior information embedding and input data adjustment, (2) model. first step, inject into network in way overcomes limitations receiving learning important contextual knowledge. And next introduce connection module between encoder decoder parts transfers more effectively from to decoder. Two public datasets, namely, OPTIMA KERMANY, are employed evaluate method. results show is an efficient vendor-independent cystoid fluid with mean Dice values 0.78 0.81 on KERMANY respectively.