From the Departments of Neurology (K.W.P., E.-J.L., S.J., M.J., J.Y.D., D.-W.K., S.J.C.) and Convergence Medicine (J.S.L., J.J., J.-G.L.), Asan Medical Center, University of Ulsan College of Medicine, Seoul; Electronics and Telecommunications Research Institute (J.S.L.), Gwangju; Promedius Inc (J.J.), Seoul; and Department of Neurology (N.C.), Heavenly Hospital, Goyang, Korea.
Neurology. 2021 Mar 30;96(13):e1761-e1769. doi: 10.1212/WNL.0000000000011654. Epub 2021 Feb 10.
We developed and investigated the feasibility of a machine learning-based automated rating for the 2 cardinal symptoms of Parkinson disease (PD): resting tremor and bradykinesia.
Using OpenPose, a deep learning-based human pose estimation program, we analyzed video clips for resting tremor and finger tapping of the bilateral upper limbs of 55 patients with PD (110 arms). Key motion parameters, including resting tremor amplitude and finger tapping speed, amplitude, and fatigue, were extracted to develop a machine learning-based automatic Unified Parkinson's Disease Rating Scale (UPDRS) rating using support vector machine (SVM) method. To evaluate the performance of this model, we calculated weighted κ and intraclass correlation coefficients (ICCs) between the model and the gold standard rating by a movement disorder specialist who is trained and certified by the Movement Disorder Society for UPDRS rating. These values were compared to weighted κ and ICC between a nontrained human rater and the gold standard rating.
For resting tremors, the SVM model showed a very good to excellent reliability range with the gold standard rating (κ 0.791; ICC 0.927), with both values higher than that of nontrained human rater (κ 0.662; ICC 0.861). For finger tapping, the SVM model showed a very good reliability range with the gold standard rating (κ 0.700 and ICC 0.793), which was comparable to that for nontrained human raters (κ 0.627; ICC 0.797).
Machine learning-based algorithms that automatically rate PD cardinal symptoms are feasible, with more accurate results than nontrained human ratings.
This study provides Class II evidence that machine learning-based automated rating of resting tremor and bradykinesia in people with PD has very good reliability compared to a rating by a movement disorder specialist.
我们开发并研究了一种基于机器学习的帕金森病（PD）两个主要症状（静止性震颤和运动迟缓）自动评分的可行性。
我们使用基于深度学习的人体姿态估计程序 OpenPose，分析了 55 名 PD 患者（110 只手臂）的静止性震颤和双侧上肢手指敲击视频片段。提取关键运动参数，包括静止性震颤幅度和手指敲击速度、幅度和疲劳度，以使用支持向量机（SVM）方法开发基于机器学习的自动统一帕金森病评定量表（UPDRS）评分。为了评估该模型的性能，我们计算了由经过运动障碍学会培训和认证的运动障碍专家对模型和金标准评分之间的加权κ和组内相关系数（ICC）。将这些值与未经训练的人类评分者和金标准评分之间的加权κ和 ICC 进行了比较。
对于静止性震颤，SVM 模型与金标准评分具有非常好到极好的可靠性范围（κ 0.791；ICC 0.927），这两个值均高于未经训练的人类评分者（κ 0.662；ICC 0.861）。对于手指敲击，SVM 模型与金标准评分具有非常好的可靠性范围（κ 0.700 和 ICC 0.793），与未经训练的人类评分者相当（κ 0.627；ICC 0.797）。
基于机器学习的算法可自动评估 PD 的主要症状，其结果比未经训练的人类评分更准确。
这项研究提供了 II 级证据，表明与运动障碍专家的评分相比，基于机器学习的 PD 患者静止性震颤和运动迟缓自动评分具有非常好的可靠性。
