Department of Biomechanics, Kinesiology and Computer Science in Sport, Centre for Sport Science and University Sports, University of Vienna, Vienna, Austria.
Research Group for Industrial Software (INSO), Vienna University of Technology, Vienna, Austria.
PLoS One. 2024 Nov 19;19(11):e0313919. doi: 10.1371/journal.pone.0313919. eCollection 2024.
The aim of this study was to determine the validity of three smartphone applications measuring barbell movement velocity in resistance training and comparing them to a commercially available linear transducer. Twenty competitive powerlifters (14 male and 6 female) completed a progressive loading protocol in the squat, bench press and deadlift (sumo or conventional) until reaching 90% of the highest load they had achieved in a recent competition. Mean velocity was concurrently recorded with three smartphone applications: Qwik VBT (QW), Metric VBT (MT), MyLift (ML), and one linear transducer: RepOne (RO). 3D motion capturing (Vicon) was used to calculate specific gold standard trajectory references for the different systems. A total of 589 repetitions were recorded with a mean velocity of (mean ± standard deviation [min-max]) 0.44 ± 0.17 [0.11-1.04] m·s-1, of which MT and ML failed to identify 52 and 175 repetitions, respectively. When compared to Vicon, RO and QW consistently delivered valid measurements (standardized mean bias [SMB] = 0 to 0.21, root mean squared error [RMSE] = 0.01 to 0.04m·s-1). MT and ML failed to deliver a level of validity comparable to RO (SMB = -0.28 to 0.14, RMSE = 0.04-0.14m·s-1), except for MT in the bench press (SMB = 0.07, RMSE = 0.04m·s-1). In conclusion, smartphone applications can be as valid as a linear transducer when assessing mean concentric barbell velocity. Out of the smartphone applications included in this investigation, QW delivered the best results.
本研究旨在确定三种智能手机应用程序在阻力训练中测量杠铃运动速度的有效性，并将其与市售的线性传感器进行比较。20 名竞技力量举重运动员（14 名男性和 6 名女性）完成了一项渐进式加载方案，包括深蹲、卧推和硬拉（相扑或传统），直到达到他们最近一次比赛中达到的最高负荷的 90%。三种智能手机应用程序（Qwik VBT[QW]、Metric VBT[MT]和 MyLift[ML]）和一个线性传感器（RepOne[RO]）同时记录平均速度。使用 3D 运动捕捉（Vicon）为不同系统计算特定的黄金标准轨迹参考。共记录了 589 次重复，平均速度为（平均值±标准差[最小值-最大值]）0.44±0.17[0.11-1.04]m·s-1，其中 MT 和 ML 分别未能识别 52 次和 175 次重复。与 Vicon 相比，RO 和 QW 始终提供有效的测量值（标准化平均偏差[SMB]为 0 至 0.21，均方根误差[RMSE]为 0.01 至 0.04m·s-1）。MT 和 ML 未能提供与 RO 相媲美的有效性水平（SMB=-0.28 至 0.14，RMSE=0.04-0.14m·s-1），除了 MT 在卧推中的表现（SMB=0.07，RMSE=0.04m·s-1）。总之，当评估平均向心杠铃速度时，智能手机应用程序可以与线性传感器一样有效。在本研究中包含的智能手机应用程序中，QW 提供了最佳的结果。
