Developing a multisensory network of wearable skin-mounted inertial measurement units to measure impulsive loads to the head

我的研究计划的长期目标是研究撞击对身体的生物力学和神经机械反应及其影响。为了实现这一目标，我的建议解决了对能够准确测量头部撞击的通用传感器系统的需求。专注于头部撞击使我能够利用我之前在评估和部署可穿戴式头部撞击传感器方面的专业知识来联系体育运动员，并为将这项技术转化为测量身体其他部位的撞击奠定基础。虽然有许多可穿戴的IMU形状系数来测量头部冲击，如嵌入在头盔中的IMU和安装在皮肤上的IMU，但由于与头骨分离，其中许多都存在显著的测量误差。对于皮肤安装的IMU传感器，头部撞击会导致底层皮肤和软组织快速振荡。安装在皮肤上的IMU记录了这些快速振荡，导致了典型的动态软组织伪影测量误差。在相对缓慢的日常运动场景中，动态软组织伪影可以被过滤掉，但在本身是突然和快速运动事件的撞击中，这是不可行的。这项拟议的研究将表征和减轻动态软组织伪影，以使用皮肤安装的IMU传感器的多传感器网络进行准确的头部撞击测量。为了实现这一点，我提出了三个短期目标，探索多感官网络方法的设计、优化和验证。首先，我们将把动态软组织伪影建模为欠阻尼的二阶系统，并描述它们与IMU传感器质量、头部位置以及与撞击方向的关系。这将使我们能够预测动态软组织伪影产生的运动学误差的预期大小，并将这些预测整合到贝叶斯估计框架中。其次，我们将优化多传感器网络中传感器的数量和位置，解决使用多传感器测量个体的实用性。最后，我们将部署多感官网络来联系体育运动员，并对多感官网络与仪表式护口器进行交叉验证，以展示与上下文相关的头部碰撞场景的准确性。这项工作包括培训8名HQP：1名博士(PhD)，3名硕士(MASC)和4名本科生(UG)，他们将从事这项研究。将这项拟议的研究转化为测量身体其他部位的影响，将增强可穿戴式皮肤安装IMU系统目前的能力，这些系统可以监测真实世界环境中的人体运动。这项技术是一个关键的组成部分，它将为研究人员提供独特的能力，准确地测量体内对人体的影响，并扩大真实世界人类运动观察的视野，以调查这些突然和快速的运动事件。

The long-term goal of my research program is to study the biomechanical and neuromechanical responses to, and effects of, impacts to the body. To achieve this goal, my proposal addresses the need for a generalizable sensor system that is capable of accurately measuring head impacts. Focusing on head impacts allows me to leverage my previous expertise in evaluating and deploying wearable head impact sensors to contact sports athletes and lays the foundation to translate the technology to measure impacts in the rest of the body. While there exist many wearable IMU form factors to measure head impacts such as IMUs embedded in headgear and IMUs mounted on skin, many of these suffer from significant measurement errors due to decoupling from the skull. For skin-mounted IMU sensors, head impacts cause the underlying skin and soft tissue to oscillate rapidly. IMUs mounted on the skin record these rapid oscillations, resulting in characteristic dynamic soft tissue artifact measurement errors. In everyday movement scenarios that are relatively slow, dynamic soft tissue artifacts can be filtered out, but in impacts that themselves are sudden and rapid movement events, this is not viable. The proposed research will characterize and mitigate dynamic soft tissue artifacts to make accurate head impact measurements using a multisensory network of skin-mounted IMU sensors. To achieve this, I propose three short-term objectives exploring the design, optimization, and validation of the multisensory network approach. First, we will model dynamic soft tissue artifacts as an under-damped second-order system and characterize their dependence on IMU sensor mass, placement on the head, and relation to the impact direction. This will allow us to predict the expected magnitude of kinematic errors produced by dynamic soft tissue artifacts and integrate these predictions in a Bayesian estimation framework. Second, we will optimize the number and placement of sensors within the multisensory network, addressing the practicality of instrumenting individuals with multiple sensors. Finally, we will deploy the multisensory network to contact sports athletes and perform a cross-validation of the multisensory network against an instrumented mouthguard, demonstrating accuracy in context-relevant head impact scenarios. This work includes the training of eight HQP: one doctoral (PhD), three Master's (MASc) and four undergraduates (UG) who will perform the research. Translation of the proposed research to measure impacts in the rest of the body will augment the current capability of wearable skin-mounted IMU systems that monitor human movement in real-world environments. This technology is a critical component that will provide researchers the unique capability of accurately measuring impacts to the body in vivo and expand the horizons of real-world human movement observations to investigate these sudden and rapid movement events.