BACKGROUND: Change of handwriting can be an early marker for severity of Parkinson's disease but suffers from poor sensitivity and specificity due to inter-subject variations. AIM: This study has investigated the group-difference in the dynamic features during sketching of spiral between PD and control subjects with the aim of developing an accurate method for diagnosing PD patients. METHOD: Dynamic handwriting features were computed for 206 specimens collected from 62 Subjects (31 Parkinson's and 31 Controls). These were analyzed based on the severity of the disease to determine group-difference. Spearman rank correlation coefficient was computed to evaluate the strength of association for the different features. RESULTS: Maximum area under ROC curve (AUC) using the dynamic features during different writing and spiral sketching tasks were in the range of 0.67 to 0.79. However, when angular features ($\boldsymbol{\varphi }$ and ${\boldsymbol{p}_{\boldsymbol{n}}}$) and count of direction inversion during sketching of the spiral were used, AUC improved to 0.933. Spearman correlation coefficient was highest for ϕ and ${\boldsymbol{p}_{\boldsymbol{n}}}$. CONCLUSION: The angular features and count of direction inversion which can be obtained in real-time while sketching the Archimedean guided spiral on a digital tablet can be used for differentiating between Parkinson's and healthy cohort.
BACKGROUND: Change of handwriting can be an early marker for severity of Parkinson's disease but suffers from poor sensitivity and specificity due to inter-subject variations. AIM: This study has investigated the group-difference in the dynamic features during sketching of spiral between PD and control subjects with the aim of developing an accurate method for diagnosing PDpatients. METHOD: Dynamic handwriting features were computed for 206 specimens collected from 62 Subjects (31 Parkinson's and 31 Controls). These were analyzed based on the severity of the disease to determine group-difference. Spearman rank correlation coefficient was computed to evaluate the strength of association for the different features. RESULTS: Maximum area under ROC curve (AUC) using the dynamic features during different writing and spiral sketching tasks were in the range of 0.67 to 0.79. However, when angular features ($\boldsymbol{\varphi }$ and ${\boldsymbol{p}_{\boldsymbol{n}}}$) and count of direction inversion during sketching of the spiral were used, AUC improved to 0.933. Spearman correlation coefficient was highest for ϕ and ${\boldsymbol{p}_{\boldsymbol{n}}}$. CONCLUSION: The angular features and count of direction inversion which can be obtained in real-time while sketching the Archimedean guided spiral on a digital tablet can be used for differentiating between Parkinson's and healthy cohort.
Authors: Julian Varghese; Catharina Marie van Alen; Michael Fujarski; Georg Stefan Schlake; Julitta Sucker; Tobias Warnecke; Christine Thomas Journal: Sensors (Basel) Date: 2021-04-30 Impact factor: 3.576