Remote Evaluation of Parkinson's Disease Using a Conventional Webcam and Artificial Intelligence

Abstract

Objective: This study aimed to prove the concept of a new optical video-based system to measure Parkinson's disease (PD) remotely using an accessible standard webcam.Methods: We consecutively enrolled a cohort of 42 patients with PD and healthy subjects (HSs). The participants were recorded performing MDS-UPDRS III bradykinesia upper limb tasks with a computer webcam. The video frames were processed using the artificial intelligence algorithms tracking the movements of the hands. The video extracted features were correlated with clinical rating using the Movement Disorder Society revision of the Unified Parkinson's Disease Rating Scale and inertial measurement units (IMUs). The developed classifiers were validated on an independent dataset.Results: We found significant differences in the motor performance of the patients with PD and HSs in all the bradykinesia upper limb motor tasks. The best performing classifiers were unilateral finger tapping and hand movement speed. The model correlated both with the IMUs for quantitative assessment of motor function and the clinical scales, hence demonstrating concurrent validity with the existing methods.Conclusions: We present here the proof-of-concept of a novel webcam-based technology to remotely detect the parkinsonian features using artificial intelligence. This method has preliminarily achieved a very high diagnostic accuracy and could be easily expanded to other disease manifestations to support PD management.

Video-Based Automated Assessment of Movement Parameters Consistent with MDS-UPDRS III in Parkinson’s Disease

“…It has been shown that remote assessment of MDS-UPDRS III is feasible via videoconferencing, except for specific items such as rigidity, or postural instability, which are not as important as bradykinesia for general evaluation of the disease [ 33, 34 ]. Indeed, some studies, which restricted bradykinesia assessment to the upper limb motor tasks of the MDS-UPDRS III, showed that upper limb motor performance was a predictive feature of PD onset and progression [ 21, 35 ]. Research on remote evaluation of PD patients has been mainly focused on the use of technological devices [ 14, 15, 36, 37 ].…”

“…Although digital tools have been recently developed in an attempt to improve bradykinesia scoring by providing quantitative measures [14][15][16], they exhibit limited use in practice because they rely on wearable sensors on patients [17,18] in a dedicated room or need specific conditions [15,[18][19][20]. More recent studies aimed to circumvent these material issues by developing video-based only tools [12,[21][22][23][24]. First, machine learning-based automatic rating for rest tremor and finger tapping was investigated to measure bradykinesia [12].…”

“…The global PD severity was assessed by applying deep learning approach to automatically score the severity level by compiling 7 of the 18 items of the MDS-UPDRS III [ 22 ]. MDS-UPDRS III bradykinesia was investigated for upper limbs tasks but with the constraint of fixed webcam recordings and with different assessment guidelines from MDS-UPDRS [ 21, 22 ]. In the same way, Guo et al [ 25 ] demonstrated that a computer-vision approach to obtain finger tapping 3D acquisition is feasible in patients with PD.…”

Video-based automated analysis of MDS-UPDRS III parameters in Parkinson disease

“…It has been shown that remote assessment of MDS-UPDRS III is feasible via videoconferencing, except for specific items such as rigidity, or postural instability, which are not as important as bradykinesia for general evaluation of the disease [30,31]. Indeed, some studies, which restricted bradykinesia assessment to the upper limb motor tasks of the MDS-UPDRS III, showed that upper limb motor performance was a predictive feature of PD onset and progression [22,32]. Research on remote evaluation of PD patients has been mainly focused on the use of technological devices [14,15,33,34].…”

“…Another work assessed the global PD severity applying deep learning approach to automatically score the severity level by compiling 7 of the 18 items of the MDS-UPDRS III [21]. Finally, Monje et al investigated MDS-UPDRS III bradykinesia upper limbs tasks but with the constraint of fixed webcam recordings and with different assessment guidelines from MDS-UPDRS [22]. Although these studies bring interesting approaches, it remains to develop tools to closely establish medically-relevant metrics for a better assessment of bradykinesia.…”