Validity of the iLOAD® app for resistance training monitoring

Abstract

Background This study aimed (I) to assess the inter-rater agreement for measuring the mean velocity (MV) of the barbell with the iLOAD® app, and (II) to compare the magnitude of the MV and total work of a training session between the iLOAD® app and a linear encoder (reference method). Method Sixteen young healthy individuals (four women and 12 men) were tested in two sessions separated by 48 h. The 10 repetition maximum (RM) load was determined in the first testing session in the half squat exercise. The second testing session consisted of 3 sets of 10 repetitions during the half squat exercise performed against the 10RM load. Both the iLOAD® app and a linear encoder were used to calculate the MV and total work of each training set. MV was recorded with the iLOAD® app by two independent researchers to evaluate the inter-rater agreement. Results Trivial differences and nearly perfect correlations were observed between raters for the MV values collected under individual sets (effect size [ES] ≤ 0.02, r ≥ 0.987), as well as for the whole training session (ES = 0.01, r = 0.997). Trivial-small differences and nearly perfect correlations were observed between the iLOAD® app and the linear encoder (Chronojump, Barcelona, Spain) for MV (EV ≤ 0.25, r ≥ 0.903) and total work (ES ≤ 0.05, r ≥ 0.973). Bland-Altman plots did not reveal heteroscedasticity of the errors between the iLOAD® app and the linear encoder for MV (r2 = 0.010) and total work (r2 < 0.001). Conclusions iLOAD® is a valid smartphone app which can provide real-time feedback of the MV and total work completed in a set of multiple repetitions in the half squat exercise.

Video-Based System for Automatic Measurement of Barbell Velocity in Back Squat

“…The Pearson’s product-moment correlation and the regression line of the measured average values and the differences between systems in a Bland–Altman plot can reveal if the systematic error is steady and independent of the sample of measured values [ 1 ]. Our results show no association between the systematic mean value and the magnitude of the random errors of mean velocity, as r 2 < 0.1, in accordance with other studies of My Lift with peak velocity values ( r 2 = 0.016 [ 17 ]) and other apps with mean values ( r 2 = 0.01 [ 47 ]). Since the random error is low and stable irrespective of the velocity range measured, the proposed video system is able to detect typical small changes in velocity needed to train and monitor high-performance athletes [ 46 ].…”

“…To that end, the proposed image processing algorithm allows for the automatic tracking of barbell markers and detection of reference points of a multipower machine to perform autocalibration in a contactless way. To the best knowledge of the authors, this is the first video-based instrument providing real-time barbell velocity outcomes without prior manual measurements of reference points in the scene [ 10 , 17 , 20 , 47 ], and without human errors due to manual video frame by frame inspection [ 10 , 17 , 20 ] or manipulation of the smartphone chronometer [ 47 ].…”

“…For the study of a smartphone app measuring mean velocity values across a series using timekeeping [ 10 ], the systematic bias is even larger and negative (–0.022 m/s), although all errors are negligible in comparison to typical velocity ranges. Likewise, the random errors given by the narrow limits of agreement observed for all the study variables (range: ± 0.47 cm, velocity: ± 0.018 m/s, force: ± 17.6 N, and power: ± 30.34 W) suggested that the video system shows less error variance than My Lift for peak (± 0.04 m/s [ 17 ], ± 0.28 m/s [ 20 ]) and mean velocity values (±0.05 m/s [ 11 ]), and other apps (± 0.034 m/s [ 47 ]). The Pearson’s product-moment correlation and the regression line of the measured average values and the differences between systems in a Bland–Altman plot can reveal if the systematic error is steady and independent of the sample of measured values [ 1 ].…”

“…The proposed video system provided valid measures of range for all variables, with practically perfect associations between systems ( r = 0.99), in accordance with My Lift assessed with mean velocity values ( r = 0.99 [ 11 ]). However, this coefficient dropped slightly when peak values were measured with the same app ( r = 0.965, 0.902, 0.963, [ 10 , 17 , 20 ], respectively) or other apps with manual timekeeping ( r = 0.948 [ 47 ]). Similarly, a low standard error of estimates (SEE) of the regression lines for all variables demonstrated very accurate predictions and low variability in the outcomes (range: 0.21 cm, velocity: 0.01 m/s, force: 8.37 N, power: 15.48 W).…”

The Implementation of Velocity-Based Training Paradigm for Team Sports: Framework, Technologies, Practical Recommendations and Challenges

“…Due to the growing interest in VBT, there has been a proliferation of devices measuring velocity for RT, from motion capture systems (MoCap), linear transducers and accelerometers, to low-cost smartphone apps. In addition, this phenomenon has been accompanied by a number of studies conducted on the validity and reliability of some of these novel technologies to measure velocity [ 21 , 22 , 23 , 24 , 25 , 26 , 27 , 28 , 29 ]. However, each of these technologies has advantages and disadvantages, and practitioners tasked with managing the systems have the responsibility to understand the pros and cons of the different systems, as well as using a systematic process in the data collection, being knowledgeable on best practices regarding implementing VBT, and applying critical thinking when assessing their RT methodologies.…”