Eletrogoniometria de punho: avaliação de sensores e procedimentos matemáticos de correção para reduzir erros de medida

Abstract

The human movement record is an essential requirement for biomechanical, clinical and occupational analysis, allowing assessments of postural variation, occupational risks and improvement of preventive programs in physical therapy and rehabilitation. Electrogoniometers have been used to record wrist movements during work. However, different sizes of sensors are used, bringing difficulties to the comparison of recordings. Moreover, although of flexible electrogoniometers (EGM) are considered a reliable and accurate equipment for dynamic recordings of different joints the EGM, they are susceptible to measurement errors, known as crosstalk. There are two main types of crosstalk, crosstalk due to sensor rotation and the inherent crosstalk. Correction procedures have been proposed to correct these errors, however no study evaluating the efficiency of corrections from both procedures in subject wrist measurements was identified in the available literature. Thus, this study aims: 1) to evaluate two different sensors in order to check its performance in an anthropometric defined sample; 2) to evaluate the effects of mathematical correction procedures on: the crosstalk due to the forearm rotation, the inherent sensors crosstalk, and the combination of these two procedures. To compare the two models of sensors 13 subjects were evaluated (7 women and 6 men) who performed flexion/extension and ulnar/radial wrist movement, reaching maximum amplitude in the pronated forearm position. Two sensors measuring 65 and 110 mm of length (XM110 and XM65, respectively) and an acquisition data device (Datalog, Biometrics) were used. The sequence of movements and the sensor placement were randomized. The data were analyzed descriptively and statistically bymultivariate analysis. The intra-individual variation (root mean square RMS between trials) were calculated. The results showed there was no significant difference between the sensors for all movements. The maximum intra individual variation was 2.2°. There were significant differences between gender for wrist flexion and ulnar deviation, women have higher values than men. The maximum variability between sensors was 7.1°. Although no significant difference was identified between sensors, it was found that the sensor XM110, bigger than XM65, presented bulging during the extension. Therefore, for individuals with height up to 1.8 m, the data obtained by both sensors are comparable. In the second stage of the study we evaluate the maximum amplitude of the flexion/extension and ulnar deviation/radial wrist of 43 healthy subjects recorded by EGMs. Correction procedures were applied to the crosstalk due to the rotation of the forearm, crosstalk inherent and combining the two procedures. The results showed there was no significant difference in measurements before and after the correction procedures (P≤0.05). Furthermore, the differences between the correction procedures were below 5 ° for most cases, indicating small clinical impact on the measurements. Considering that the corrections are time consuming, require specific technical knowledge and have provided inefficient results, the correction procedures are not recommended for clinical wrist electrogoniometers recordings. Thus, improvement of equipment should be provided by the manufacturers.