Draw4CM: Detecting Cervical Myelopathy via Hand Drawings Captured by Mobile Devices

Cervical myelopathy (CM), which causes numbness and pain in the hands, can affect manual dexterity and render daily life difficult. The characteristics of this disease are expected to be apparent owing to the necessity of fine motor control for drawing in daily life. Therefore, this study proposed a method to screen for CM by asking patients to draw a figure on a tablet screen using a stylus. The drawing was captured as an image and fed into our machine-learning model to predict CM. The proposed models exhibited sensitivity and specificity comparable to or better than conventional physical methods while avoiding examiner subjectivity and bias. Further, we developed a smartphone application that uses the camera to capture images of figures drawn on paper for CM screening. Moreover, the proposed methods can enable smartphones and tablets to capture disease characteristics and are ubiquitous, thus rendering it easier to reach potential patients.