Assessment and evaluation of Hill-type muscle models for predicting in vivo force

用于预测体内力的 Hill 型肌肉模型的评估和评价

• 批准号: 7927041
• 负责人: Andrew A Biewener
• 金额: $ 50.45万
• 依托单位: HARVARD UNIVERSITY
• 依托单位国家: 美国
• 财政年份: 2008
• 资助国家: 美国
• 起止时间: 2008-09-26 至 2012-08-31
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/7927041
• 关键词: Address; Affect; Aging; Algorithms; Animal Model; Animals; Architecture; Basic Science; Behavior; Biomechanics; Characteristics; Clinical; Clinical assessments; Data; Development; Distal; Electric Stimulation; Electrodes; Evaluation; Exercise; Fascicle; Felis catus; Fiber; Frequencies; Gait; Goals; Goat; Human; In Situ; Individual; Laboratories; Length; Life; Limb structure; Locomotion; Measurement; Measures; Methods; Modeling; Motor; Motor output; Muscle; Muscle Fibers; Muscle function; Outcome; Output; Patients; Pattern; Performance; Physical activity; Physiological; Population; Process; Property; Prosthesis; Public Health; Recruitment Activity; Rehabilitation therapy; Relative (related person); Research; Research Personnel; Signal Transduction; Simulate; Skeletal Muscle; Skin; Stroke; Surface; Techniques; Tendon force; Testing; Therapeutic Intervention; Time; Translating; Wheelchairs; Work; base; design; grasp; improved; in vivo; motor disorder; motor impairment; neuromuscular function; novel; orthotics; public health relevance; relating to nervous system; spatiotemporal

DESCRIPTION (provided by applicant): The overarching goal of the proposed research is to improve the quality of understanding and assessment of neuromotor performance that can be obtained through the use of electromyographic (EMG) recordings of muscle activity patterns and Hill-type muscle models. Muscle modeling and EMG analysis has widespread use for improving the assessment and development of rehabilitation therapies important to the treatment of motor impairment, as well as changes in muscle function associated with aging. EMG recordings, whether from indwelling electrodes or measured from the skin surface, are frequently used in combination with muscle models to simulate and evaluate motor performance to address a broad range of clinical problems and therapies that include gait rehabilitation, the evaluation and treatment of stroke, wheel chair use, and prosthetics. This work seeks to combine cutting-edge basic science analysis of muscle properties and in vivo contractile function with computational muscle models for interpreting the contractile performance of whole muscles relative to their motor recruitment patterns. The proposed work is designed to directly test and refine the models, facilitating improvements to the quality of muscle modeling that can be applied in human neuromotor studies to a range of clinical problems and conditions. By combining direct in vivo recordings of muscle force (via tendon force buckles), fascicle length change (via sonomicrometry), and neural activation (via multiple indwelling fine-wire EMG electrodes) in an animal model (goat hind limb muscles), quantitative measures of in vivo contractile performance will be used to validate and improve the fit of four different Hill-type muscle models based on muscle activation and architecture. Spatio-temporal features of the EMG signals recorded within the muscles will be analyzed using wavelets to examine patterns of motor recruitment in relation to in vivo contractile performance of select muscles. These will be used to derive and test activation patterns used as input to the muscle models. Fundamental features, such as the Henneman size-principle for orderly recruitment and changes in work output (concentric versus eccentric exercise), will be examined to test and refine the models. Sensitivity analyses will also be carried out to test model output robustness against known changes in model input parameters derived from in situ muscle measurements of activation and force development rates, F-L properties, and Vmax. The following two specific aims will be examined: Aim #1 will examine the ability of different Hill-type muscle models to characterize measured patterns of whole muscle force and work output under in vivo conditions, based on activation input derived from the fine-wire EMG recordings. Time-frequency spectra of the EMGs will be analyzed to reveal patterns of motor unit recruitment, testing the hypotheses that: (a) differential patterns of motor recruitment (between the fast and slow units) occur during goat locomotion, and (b) the faster motor units are preferentially activated, relative to slow units, for tasks that require high strain rates and high rates of force development. Measurements of intrinsic in situ muscle properties, architecture and fiber type will also test the hypothesis that a homogeneous distribution of fiber types and pennation angle within muscle regions results in uniform patterns of fascicle strain and contractile function for a given type of locomotor behavior. Aim #2 will analyze detailed spatio-temporal features of the EMG recordings made within local muscle regions of select limb muscles using wavelets to provide a quantitative time-varying evaluation of motor unit recruitment. In situ recordings of twitch force development and slack-test releases will provide estimates of the intrinsic properties of the different motor units. Wavelet analysis will be used to refine and improve algorithms developed for the activation/deactivation dynamics used in the muscle models to improve their fit to direct measurements of muscle contractile performance. Aim #2 will test the hypothesis that the time-varying patterns of whole muscle force development are better predicted by muscle model that incorporate the actual in vivo motor recruitment patterns tha models that do not. PUBLIC HEALTH RELEVANCE: The relevance of the proposed research to public health is that it will help improve the clinical assessment of neuromotor performance that can be obtained through non-invasive use of electromyographic (EMG) recordings of patient muscle activity patterns associated with particular motor functions, such as gait or manipulation and grasping. EMG recordings are commonly made from surface (skin mounted) electrodes to assess neuromuscular function in an individual. These muscle activity recordings are then interpreted to assess and develop rehabilitation therapies, important to the treatment of motor impairment, such as that which results from stroke, as well as changes in muscle function associated with aging. Muscle researchers also widely use Hill-type muscle models derived from known physiological force-velocity and force-length properties of skeletal muscle to simulate or predict the motor output of a muscle based on its measured EMG activation. The combination of non-invasive EMG recordings as input to drive muscle models for predicting biomechanical outcomes is frequently used to address a broad range of clinical problems and therapies, including functional electrical stimulation, applied to gait rehabilitation, the evaluation and treatment of stroke, and prosthetics and orthotics. However, in humans, such models of an individual's muscles cannot be tested directly. Further, most muscle models assume uniform motor unit characteristics, whereas most muscles have mixed populations of motor units that can be differentially recruited. Consequently, the proposed work seeks to combine cutting-edge basic science analysis of muscle properties and in vivo contractile function, based on novel recording and analysis methods, with computational muscle models that will allow the models' output to be assessed directly by the measurements of the muscle's contractile performance in the living animal. Goat muscle function will be assessed across a range of physical activity using methods that allow muscle force, length change, and activation to be recorded in vivo. Wavelet decomposition of regional EMGs within the muscle will allow the recruitment patterns of motor units to be identified in relation to changes in contractile performance. The proposed work is designed to facilitate the refinement of Hill-type muscle models to improve their ability to predict muscle force and work output that can be obtained from non-invasive EMG recordings of muscle, which are commonly made in the clinical laboratory setting and applied to the assessment and treatment of broad range of motor disorders and conditions.

描述（由申请人提供）：拟议研究的总体目标是提高对神经运动性能的理解和评估的质量，神经运动性能可以通过使用肌肉活动模式的肌电图（EMG）记录和希尔型肌肉模型来获得。肌肉建模和肌电图分析广泛用于改善康复疗法的评估和开发，这对于治疗运动障碍以及与衰老相关的肌肉功能变化非常重要。肌电图记录，无论是来自留置电极还是从皮肤表面测量，经常与肌肉模型结合使用来模拟和评估运动表现，以解决广泛的临床问题和治疗，包括步态康复、中风的评估和治疗、轮椅的使用和假肢。这项工作旨在将肌肉特性和体内收缩功能的尖端基础科学分析与计算肌肉模型结合起来，以解释整个肌肉相对于其运动募集模式的收缩性能。拟议的工作旨在直接测试和完善模型，促进提高肌肉建模的质量，该模型可应用于人类神经运动研究中的一系列临床问题和条件。通过结合动物模型（山羊后肢肌肉）中肌肉力（通过肌腱力扣）、肌束长度变化（通过声测法）和神经激活（通过多个留置细线肌电图电极）的直接体内记录，定量测量体内收缩性能的测量将用于验证和改进基于肌肉激活和架构的四种不同 Hill 型肌肉模型的拟合度。将使用小波分析肌肉内记录的肌电图信号的时空特征，以检查与选定肌肉的体内收缩性能相关的运动募集模式。这些将用于导出和测试用作肌肉模型输入的激活模式。将检查基本特征，例如用于有序招募的亨尼曼尺寸原则和工作输出的变化（同心与偏心运动），以测试和完善模型。还将进行敏感性分析，以测试模型输出的鲁棒性，以对抗模型输入参数的已知变化，这些变化源自原位肌肉测量的激活和力发展速率、F-L 特性和 Vmax。将检查以下两个具体目标： 目标#1 将检查不同 Hill 型肌肉模型的能力，以根据细线肌电图导出的激活输入来表征体内条件下整个肌肉力量和工作输出的测量模式。录音。将分析肌电图的时间频谱，以揭示运动单位募集的模式，测试以下假设：（a）山羊运动过程中出现运动募集的差异模式（快速和慢速单位之间），以及（b）速度越快对于需要高应变率和高发力率的任务，相对于慢速单位，运动单位优先被激活。内在原位肌肉特性、结构和纤维类型的测量还将检验这样的假设：肌肉区域内纤维类型和羽状角度的均匀分布导致给定类型的运动行为的束应变和收缩功能的均匀模式。目标#2将使用小波分析在选定肢体肌肉的局部肌肉区域内进行的肌电图记录的详细时空特征，以提供运动单位募集的定量时变评估。抽搐力发展和松弛测试释放的现场记录将提供对不同运动单位的内在特性的估计。小波分析将用于细化和改进为肌肉模型中使用的激活/失活动力学开发的算法，以提高其对肌肉收缩性能直接测量的拟合度。目标 #2 将检验这样的假设：与不包含实际体内运动募集模式的肌肉模型相比，包含实际体内运动募集模式的肌肉模型可以更好地预测整个肌肉力量发展的时变模式。公共健康相关性：拟议研究与公共健康的相关性在于，它将有助于改善神经运动性能的临床评估，该评估可以通过无创地使用与特定运动功能相关的患者肌肉活动模式的肌电图（EMG）记录来获得，例如步态或操纵和抓握。 EMG 记录通常由表面（安装在皮肤上）电极进行，以评估个体的神经肌肉功能。然后对这些肌肉活动记录进行解释，以评估和开发康复疗法，这对于治疗运动损伤（例如中风引起的运动损伤）以及与衰老相关的肌肉功能变化很重要。肌肉研究人员还广泛使用源自骨骼肌已知生理力速度和力长度特性的 Hill 型肌肉模型，根据测量的肌电图激活来模拟或预测肌肉的运动输出。非侵入性肌电图记录的组合作为驱动肌肉模型的输入，以预测生物力学结果，经常用于解决广泛的临床问题和治疗，包括功能性电刺激，应用于步态康复、中风的评估和治疗，以及假肢和矫形器。然而，在人类中，这种个人肌肉模型无法直接测试。此外，大多数肌肉模型假设统一的运动单位特征，而大多数肌肉具有可以差异性招募的混合运动单位群。因此，拟议的工作旨在将基于新颖的记录和分析方法的肌肉特性和体内收缩功能的尖端基础科学分析与计算肌肉模型结合起来，该模型将允许通过测量直接评估模型的输出活体动物肌肉的收缩性能。将使用允许在体内记录肌肉力量、长度变化和激活的方法来评估一系列身体活动中的山羊肌肉功能。肌肉内区域肌电图的小波分解将允许识别与收缩性能变化相关的运动单位的募集模式。拟议的工作旨在促进 Hill 型肌肉模型的细化，以提高其预测肌肉力量和工作输出的能力，这些肌肉力量和工作输出可以从肌肉的非侵入性肌电图记录中获得，这些记录通常在临床实验室环境中进行并应用评估和治疗广泛的运动障碍和病症。