Fig. 1From: Automating Rey Complex Figure Test scoring using a deep learning-based approach: a potential large-scale screening tool for cognitive declineRCFT images pre-processing. First, noise filtering and binarization were applied (step 1). Second, we rotated the images to correct skewness in the process of scanning correction and removed all unrelated background images such as the subject’s name or the number of page (step 2). Third, we obtained the contours of each image to crop the images’ bounding rectangles, and finally, all images were resized to \(512\times 512\) pixels (step 3)Back to article page