Alagha M Abdulhadi, Alagha Mahmoud A, Dunstan Eleanor, Sperwer Olaf, Timmins Kate A, Boszczyk Bronek M
Academic Orthopaedics, Trauma and Sports Medicine, Queen's Medical Centre, The University of Nottingham, Nottingham, NG7 2UH, UK.
Medical, Veterinary and Life Sciences College, University of Glasgow, Dental Hospital, Glasgow, G2 3JZ, UK.
Eur Spine J. 2017 Apr;26(4):1291-1297. doi: 10.1007/s00586-017-4948-3. Epub 2017 Jan 19.
To assess the reliability and validity of a hand motion sensor, Leap Motion Controller (LMC), in the 15-s hand grip-and-release test, as compared against human inspection of an external digital camera recording.
Fifty healthy participants were asked to fully grip-and-release their dominant hand as rapidly as possible for two trials with a 10-min rest in-between, while wearing a non-metal wrist splint. Each test lasted for 15 s, and a digital camera was used to film the anterolateral side of the hand on the first test. Three assessors counted the frequency of grip-and-release (G-R) cycles independently and in a blinded fashion. The average mean of the three was compared with that measured by LMC using the Bland-Altman method. Test-retest reliability was examined by comparing the two 15-s tests.
The mean number of G-R cycles recorded was: 47.8 ± 6.4 (test 1, video observer); 47.7 ± 6.5 (test 1, LMC); and 50.2 ± 6.5 (test 2, LMC). Bland-Altman indicated good agreement, with a low bias (0.15 cycles) and narrow limits of agreement. The ICC showed high inter-rater agreement and the coefficient of repeatability for the number of cycles was ±5.393, with a mean bias of 3.63.
LMC appears to be valid and reliable in the 15-s grip-and-release test. This serves as a first step towards the development of an objective myelopathy assessment device and platform for the assessment of neuromotor hand function in general. Further assessment in a clinical setting and to gauge healthy benchmark values is warranted.
评估手部运动传感器Leap Motion Controller（LMC）在15秒握放测试中的可靠性和有效性，并与通过外部数码相机记录进行人工检查的结果进行比较。
50名健康参与者被要求尽可能快速地完全握放其优势手，进行两次测试，中间休息10分钟，同时佩戴非金属手腕夹板。每次测试持续15秒，第一次测试时使用数码相机拍摄手部的前外侧。三名评估人员独立且盲法计数握放（G-R）周期的频率。使用Bland-Altman方法将三人的平均值与LMC测量的值进行比较。通过比较两次15秒测试来检验重测可靠性。
记录的G-R周期平均数为：47.8±6.4（测试1，视频观察者）；47.7±6.5（测试1，LMC）；以及50.2±6.5（测试2，LMC）。Bland-Altman分析表明一致性良好，偏差低（0.15个周期）且一致性界限窄。组内相关系数显示评估者间一致性高，周期数的重复性系数为±5.393，平均偏差为3.63。
LMC在15秒握放测试中似乎是有效且可靠的。这是朝着开发用于评估一般神经运动手部功能的客观脊髓病评估设备和平台迈出的第一步。有必要在临床环境中进行进一步评估并确定健康基准值。
