Automatic Classification System for Diagnosis of Cognitive Impairment Based on the Clock-Drawing Test

The prevalence of dementia is currently increasing worldwide. This syndrome produces a deterioration in cognitive function that can not be reverted. However, an early diagnosis can be crucial for slowing its progress. The Clock Drawing Test (CDT) is a widely used paperand-pencil test for cognitive assessment in which an individual has to manually draw a clock on a paper during a certain time. Nevertheless, there are a lot of scoring systems for this test and most of them depend on the subjective assessment of the expert. This study proposes a computeraided diagnosis (CAD) system based on deep learning in order to automate the diagnosis of cognitive impairment (CI) from the result of the CDT. This is addressed by employing a preprocessing pipeline in which the clock is detected and centered, as well as binarized for decreasing the computational burden. Then, the resulting image is fed into a Convolutional Neural Network (CNN), which is used to identify the informative patterns within the CDT drawings that are relevant for the assessment of the patient's cognitive status. Performance is evaluated in a real context where differentiating between CI patients and controls. The proposed method provides an accuracy of 68.62% in this classification task, with an AUC of 74.53%. A validation method using resubstitution with upper bound correction is also discussed.