Development of a Drawing Application to Evaluate Hand and Wrist Function: A Pilot Study.

Introduction: We developed a custom digital drawing application to assess hand function. We conducted an initial validation study of this technique to (1) assess which drawing features are associated with hand function, (2) differentiate patients from control subjects for both dominant and nondominant hands, and (3) assess the correlation of drawing features with previously validated patient-reported outcome measures (PROMs).

Methods: In this prospective study, participants were asked to draw shapes on an Apple iPad with a digital pen using a custom app. Drawings from 142 hands in 73 participants were categorized based on hand dominance and patient/control subject. We calculated kinematic/geometric and pressure-based features from raw drawing data. Random forest models were used to classify patients and control subjects and to identify correlation with validated PROMs. Model performance for classification was assessed using accuracy, precision, recall, F1 score, and area under the curve.

Results: Patients and control subjects could not be differentiated by visual inspection; however, many drawing features were different (P < 0.05) between patients and control subjects for both dominant (78 features) and nondominant (27 features) hand drawings. Circle drawings were most informative, and pressure features were most important. The classification metrics including area under the curve (0.82 to 0.84), accuracy (0.75 to 77), and F1 score (0.78 to 0.81) suggest that hand function is reasonably assessed through drawing. Drawing features were correlated with patient-rated wrist evaluation, 12-Item Short Form Health Survey, and Quick Disabilities of the Arm, Shoulder and Hand scores (P < 0.001).

Discussion: We developed a new technique to objectively measure hand function using drawing. Drawing features were correlated with validated PROMs and could differentiate patients from control subjects, regardless of hand dominance.