Leiden University Medical Centre, Department of Neurology, The Netherlands.
Leiden Institute of Advanced Computer Science, The Netherlands.
Clin Neurophysiol. 2023 Feb;146:49-54. doi: 10.1016/j.clinph.2022.11.019. Epub 2022 Dec 9.
Distinguishing normal, neuropathic and myopathic electromyography (EMG) traces can be challenging. We aimed to create an automated time series classification algorithm.
EMGs of healthy controls (HC, n = 25), patients with amyotrophic lateral sclerosis (ALS, n = 20) and inclusion body myositis (IBM, n = 20), were retrospectively selected based on longitudinal clinical follow-up data (ALS and HC) or muscle biopsy (IBM). A machine learning pipeline was applied based on 5-second EMG fragments of each muscle. Diagnostic yield expressed as area under the curve (AUC) of a receiver-operator characteristics curve, accuracy, sensitivity, and specificity were determined per muscle (muscle-level) and per patient (patient-level).
Diagnostic yield of the classification ALS vs. HC was: AUC 0.834 ± 0.014 at muscle-level and 0.856 ± 0.009 at patient-level. For the classification HC vs. IBM, AUC was 0.744 ± 0.043 at muscle-level and 0.735 ± 0.029 at patient-level. For the classification ALS vs. IBM, AUC was 0.569 ± 0.024 at muscle-level and 0.689 ± 0.035 at patient-level.
An automated time series classification algorithm can distinguish EMGs from healthy individuals from those of patients with ALS with a high diagnostic yield. Using longer EMG fragments with different levels of muscle activation may improve performance.
In the future, machine learning algorithms may help improve the diagnostic accuracy of EMG examinations.
区分正常、神经源性和肌源性肌电图（EMG）记录可能具有挑战性。我们旨在创建一种自动化时间序列分类算法。
根据纵向临床随访数据（ALS 和 HC）或肌肉活检（IBM），回顾性选择健康对照者（HC，n=25）、肌萎缩侧索硬化症（ALS，n=20）和包涵体肌炎（IBM，n=20）的 EMG。基于每个肌肉的 5 秒 EMG 片段应用机器学习管道。通过接收者操作特征曲线的曲线下面积（AUC）、准确性、敏感性和特异性来确定每个肌肉（肌肉水平）和每个患者（患者水平）的诊断效果。
分类 ALS 与 HC 的诊断效果为：肌肉水平 AUC 为 0.834±0.014，患者水平 AUC 为 0.856±0.009。分类 HC 与 IBM 的 AUC 为肌肉水平 0.744±0.043，患者水平 0.735±0.029。分类 ALS 与 IBM 的 AUC 为肌肉水平 0.569±0.024，患者水平 0.689±0.035。
自动化时间序列分类算法可以区分健康个体和 ALS 患者的 EMG，具有较高的诊断效果。使用不同肌肉激活水平的更长 EMG 片段可能会提高性能。
未来，机器学习算法可能有助于提高 EMG 检查的诊断准确性。
