利用雙陀螺儀作下肢開放式動力鏈運動分類

Classifying open kinetic chain movements in lower extremities by two gyros

背景和目的：過去許多研究探討無線慣性感測器用於人類活動和日常生活的應用，但很少有人討論過它們在動作中的追蹤監測功能。本研究的重點是利用慣性感測器記錄腿部動作搭配簡單演算法，對腿部動作的運動數據進行分類。

方法：將兩個慣性感測元件固定於受試者的膝關節和外踝搭配遠程監控系統以測量其開放動力鏈運動數據，藉由分析兩個不同位置的慣性感測元件的同步角速資料將運動開放動力鏈運動作分類。

結果：此元件組和運算法對膝關節屈曲/伸展（99.8％），髖關節屈曲（95.8％）和膝關節內旋/外旋（85.3％）的敏感度非常高，但髖關節外展/內收（38.3％）的敏感度偏低。然而，所建立的演算法對四種類型的運動均具有良好的特異性(80.5％-99.8％)。

結論：適當地固定慣性感測元件配合演算法可成為一個理想穿戴是傳感器，此動作追蹤工具可作為遠距復健的良好監測工具。

 

Background and purpose: A number of studies have examined the applicability of inertial sensors to human motion tracking and daily activities, but few have discussed their application to monitor exercises. This study focuses on leg motion tracking and the development of a simple kinematic based algorithm to classify leg motion data. Methods: A tele-monitoring system is proposed, using two inertial measurement units (IMU) are applied to each participant’s knee and lateral malleolus to measure their open kinetic chain motion data. The motion classification was done through interpreting angular velocities simultaneously from two different IMUs. Results: There was a good sensitivity for knee flexion/extension (99.8%), hip flexion (95.8%) and knee internal rotation/external rotation (85.3%) and very low sensitivity for hip abduction/adduction (38.3%). However, the established algorithm owns good specificity for the four types of exercise, ranging from 80.5% to 99.8%. The causes of misinterpretation in this study were also discussed. Conclusions: The results show that properly affixed inertial measurement units and a suitable algorithm provide ideal wearable sensors for telerehabilitation.