List of fit indices when fitting a linear LCM.

We focus on fine vibrations originating from tendons (Mechanotendography: MTG) as a novel method for quantifying muscle activity. Quantifying muscle activity using MTG can enable daily and long-term continuous measurements, which have been challenging for electromyography (EMG) and mechanomyography (MMG). However, the detailed trajectory of MTG increase relative to exerted muscle strength has not been clarified, nor has the mechanism of MTG generation. Our research has two objectives. The first is to clarify the detailed relationship between exerted muscle strength levels and MTG through statistical modeling. The second is to establish a highly accurate hypothesis concerning the mechanism of MTG generation based on the modeling results and physiological knowledge. We focused on the Achilles tendon to study these two objectives. Experiments were conducted on 62 participants, and MTG data were obtained at various levels of exerted muscle strength. The obtained data were structured into a longitudinal data format representing the trajectory of MTG increase with increasing exerted muscle strength. We used latent curve models (LCM) to identify this structure. By applying various LCMs to explore an optimal model, we found that the quadratic LCM received the best fit for females, while the piecewise linear LCM with a breakpoint at 50% exerted muscle strength received the best fit for males. Notably, a significant sex difference was observed in the rate of increase in MTG at low levels of exerted muscle strength. These results suggest that MTG is caused by fine vibrations generated by muscle fiber contractions, and these fine vibrations are transmitted to the tendons connected to the muscles, where they are observed. Future research will focus on verifying this hypothesis through increased time points and physiological experiments.本研究以源自肌腱的细微振动对应的肌腱振动法（Mechanotendography, MTG）作为量化肌肉活动的创新手段。相较于肌电图（electromyography, EMG）与肌机械图（mechanomyography, MMG），采用MTG量化肌肉活动可实现日常化与长期连续测量，而这两类传统方法此前难以达成此类测量目标。然而，MTG信号增幅与肌肉施力强度间的详细变化轨迹尚未明确，MTG的产生机制亦未得到阐明。本研究设有两项核心目标：其一为通过统计建模，阐明肌肉施力强度水平与MTG信号间的具体关联；其二为基于建模结果与生理学知识，构建关于MTG产生机制的高精度假说。为开展上述两项研究，本研究选取跟腱作为研究靶点。研究共招募62名受试者，采集了不同施力强度水平下的MTG数据。所获数据被整理为纵向数据格式，用以表征MTG信号随肌肉施力强度提升而增长的变化轨迹。本研究采用潜曲线模型（latent curve models, LCM）对该数据结构进行识别。通过拟合多种LCM以探索最优模型，结果显示：二次型LCM对女性受试者数据适配效果最优，而以50%施力强度为断点的分段线性LCM对男性受试者数据适配效果最佳。值得注意的是，在低施力强度区间内，MTG信号增幅速率存在显著的性别差异。上述结果表明，MTG信号源自肌纤维收缩产生的细微振动，此类振动经传递至与肌肉相连的肌腱后可被检测到。未来研究将通过增加时间采样点与开展生理学实验，对本次提出的假说进行验证。