Verification of reliability of measurement systems for swimming analysis

Abstract

Sophisticated technologies and measuring devices are increasingly being used for the analysis of performance and swimming techniques. At BUT, the Tachograph II and Swimming Inertial Measure Unit (SwIMU) measuring systems were developed for swimming analysis. To evaluate the measured data, the SwimDataViewer software was developed for both measuring systems. To verify the reliability of the Tachograph, 34 race swimmers were measured (15 F, 19 M), and 28 race swimmers (16 M, 12 F) were measured with SwIMU. All swimmers swam a 6x25m front crawl. Pullout and the first three and last two strokes of the arms were not included in the data evaluation. The intraclass correlation coefficient (ICC) was used to verify the reliability of the monitored parameters. The values of measured parameters in the front crawl for Tachograph (SwIMU) were ICC = 0.992 (0.991) for the average value of speed, ICC = 0.958 (0.960) for efficiency, ICC = 0.971 (0.988) for the frequency of the arms, and finally, ICC = 0.978 (0.992) for the frequency of the legs. The ICC coefficient for all monitored parameters in both devices showed a high degree of reliability. From this perspective, both devices are suitable for use in coaching and research practice.Sophisticated technologies and measuring devices are increasingly being used for the analysis of performance and swimming techniques. At BUT, the Tachograph II and Swimming Inertial Measure Unit (SwIMU) measuring systems were developed for swimming analysis. To evaluate the measured data, the SwimDataViewer software was developed for both measuring systems. To verify the reliability of the Tachograph, 34 race swimmers were measured (15 F, 19 M), and 28 race swimmers (16 M, 12 F) were measured with SwIMU. All swimmers swam a 6x25m front crawl. Pullout and the first three and last two strokes of the arms were not included in the data evaluation. The intraclass correlation coefficient (ICC) was used to verify the reliability of the monitored parameters. The values of measured parameters in the front crawl for Tachograph (SwIMU) were ICC = 0.992 (0.991) for the average value of speed, ICC = 0.958 (0.960) for efficiency, ICC = 0.971 (0.988) for the frequency of the arms, and finally, ICC = 0.978 (0.992) for the frequency of the legs. The ICC coefficient for all monitored parameters in both devices showed a high degree of reliability. From this perspective, both devices are suitable for use in coaching and research practice.