Assessment and Evaluation of Hill-type Muscle Models for Predicting In Vivo Force

描述(申请人提供)：Hill-type肌肉模型广泛适用于运动功能的评估，对改进的康复策略的设计至关重要，是肌肉驱动模拟的关键组成部分，旨在确定由于年龄或神经肌肉损伤而限制行动能力的因素。尽管希尔型模型的使用无处不在，但很少有研究在体内、时变的条件下检验它们的准确性和有效性。拟议项目的主要目标是测试和改进使用先进的Hill类型模型评估人类肌肉功能的方法，以及肌电活动和肌肉结构的非侵入性测量。主要的创新目标是在肌肉驱动的人类自行车模拟中实现我们最近开发的具有独立快和慢收缩元素的Hill类型模型，并在一系列速度和测功条件下预测腓肠肌力，与基于三维超声的肌腱应变测量实验确定的力相比，预测效果更好。我们之前的工作表明，由来自肌电记录的快、慢运动单位的招募模式驱动的新的双元素模型，比具有平均纤维特性的传统的单元素模型对原位和活体肌力的预测要好得多。然而，由于我们的山羊动物模型太大，我们无法从实验上评估肌肉的F-V特性，这可能降低了模型的预测能力。这项拟议的研究通过使用一种新的小动物模型(大鼠后肢远端肌肉)来解决这一局限性，以创新的现场分析不同刺激条件下的F-V和循环功率输出。这些分析，加上对肌肉形状变化的活体3D X射线成像，将进一步推动两元素模型的发展。目前的工作针对两个特定的目标，关键地扩大了Hill类型模型对与康复相关的人体运动功能的临床评估的影响：目标1评估了单元素模型和两元素模型在特定对象的自行车测功器上骑行的特定对象模拟中估计随时间变化的肌力的准确性，使用基于新型3D超声的肌腱应变、束状肌和肌肉厚度的测量。目的#2研究运动单位募集和刺激频率如何影响原位肌肉的机械输出，目的是更好地预测肌力和在体及在体的力量。两元素模型的改进将基于原位收缩动力学、新的在体肌肉-肌腱力和肌束应变测量方法，以及创新的3D X射线视频透视。来自目标2的见解将被迭代地纳入在目标1中测试的人类自行车模拟中。

DESCRIPTION (provided by applicant): Hill-type muscle models are broadly applicable to the assessment of motor function and critical to the design of improved rehabilitative strategies, serving as key components of muscle- driven simulations aimed at identifying factors that limit mobility due to age or neuromuscular impairment. Despite the ubiquitous use of Hill-type models, few studies have examined their accuracy and validity under in vivo, time-varying conditions. An overarching goal of the proposed project is to test and refine methods for assessing human muscle function using advanced Hill-type models, together with non-invasive measures of electromyographic activity and muscle structure. The lead innovative aim is to implement our recently developed two-element Hill-type model, with independent fast and slow contractile elements, within muscle-driven simulations of human cycling and to predict gastrocnemius forces across a range of speed and ergometry conditions that compare favorably to the forces determined experimentally from 3D ultrasound-based measures of tendon strain. Our previous work showed that a novel two-element model, driven by recruitment patterns of fast and slow motor units derived from EMG recordings, generates significantly better predictions of in situ and in vivo muscle force than traditional one-element models with average fiber properties. However, due to the large size of our goat animal model, we were unable to experimentally assess the muscles' F-V characteristics, which may have diminished the model's predictive capability. The proposed research addresses this limitation by using a new small animal model (rat distal hindlimb muscles) to conduct innovative in situ analyses of F-V and cyclical power output under varying stimulation conditions. These analyses, together with in vivo 3D X-ray imaging of muscle shape changes, will further advance the two-element model. The current work addresses two specific aims, critically broadening the impact of Hill-type models on clinical assessment of human motor function related to rehabilitation: Aim #1 evaluates the accuracy with which one- vs two-element models can estimate time-varying muscle forces within subject-specific simulations of human subjects pedaling on a cycle ergometer, using novel 3D ultrasound-based measures of tendon strain, fascicle pennation and muscle thickness. Aim #2 examines how motor unit recruitment and stimulation frequency affect in situ muscle mechanical output, with the goal of better predicting muscle force and power in situ and in vivo. Refinement of the two-element model will be based on in situ contractile dynamics, novel in vivo muscle-tendon force and fascicle strain measures, and innovative 3D X-ray video fluoroscopy. Insights from Aim 2 will be iteratively incorporated into the simulations of human cycling tested in Aim 1.