Department of Neurosurgery, Medical College of Wisconsin, Milwaukee, WI 53226, USA.
Injury Biomechanics and Protection Group, U.S. Army Aeromedical Research Laboratory, Fort Rucker, AL 36362, USA.
Mil Med. 2023 Nov 8;188(Suppl 6):420-427. doi: 10.1093/milmed/usad178.
Because brain regions are responsible for specific functions, regional damage may cause specific, predictable symptoms. However, the existing brain injury criteria focus on whole brain response. This study developed and validated a detailed human brain computational model with sufficient fidelity to include regional components and demonstrate its feasibility to obtain region-specific brain strains under selected loading.
Model development used the Simulated Injury Monitor (SIMon) model as a baseline. Each SIMon solid element was split into 8, with each shell element split into 4. Anatomical regions were identified from FreeSurfer fsaverage neuroimaging template. Material properties were obtained from literature. The model was validated against experimental intracranial pressure, brain-skull displacement, and brain strain data. Model simulations used data from laboratory experiments with a rigid arm pendulum striking a helmeted head-neck system. Data from impact tests (6 m/s) at 2 helmet sites (front and left) were used.
Model validation showed good agreement with intracranial pressure response, fair to good agreement with brain-skull displacement, and good agreement for brain strain. CORrelation Analysis scores were between 0.72 and 0.93 for both maximum principal strain (MPS) and shear strain. For frontal impacts, regional MPS was between 0.14 and 0.36 (average of left and right hemispheres). For lateral impacts, MPS was between 0.20 and 0.48 (left hemisphere) and between 0.22 and 0.51 (right hemisphere). For frontal impacts, regional cumulative strain damage measure (CSDM20) was between 0.01 and 0.87. For lateral impacts, CSDM20 was between 0.36 and 0.99 (left hemisphere) and between 0.09 and 0.93 (right hemisphere).
Recognizing that neural functions are related to anatomical structures and most model-based injury metrics focus on whole brain response, this study developed an anatomically accurate human brain model to capture regional responses. Model validation was comparable with current models. The model provided sufficient anatomical detail to describe brain regional responses under different impact conditions.
由于大脑区域负责特定的功能，区域损伤可能导致特定的、可预测的症状。然而，现有的脑损伤标准侧重于整个大脑的反应。本研究开发并验证了一个详细的人类大脑计算模型，该模型具有足够的逼真度，可以包括区域成分，并证明在选定的加载下获得特定区域脑应变的可行性。
模型开发使用 Simulated Injury Monitor (SIMon) 模型作为基线。每个 SIMon 实体单元被细分为 8 个，每个壳单元被细分为 4 个。从 FreeSurfer fsaverage 神经影像学模板中识别出解剖区域。材料特性取自文献。该模型通过实验颅内压、脑颅骨位移和脑应变数据进行验证。模型模拟使用刚性臂摆锤撞击头盔式头颈部系统的实验室实验数据。使用 2 个头盔部位（前侧和左侧）的 6m/s 冲击测试数据。
模型验证表明，颅内压响应吻合良好，脑颅骨位移吻合良好至较好，脑应变吻合良好。最大主应变 (MPS) 和剪应变的相关分析分数均在 0.72 到 0.93 之间。对于前向冲击，区域 MPS 在 0.14 到 0.36 之间（左右半球平均值）。对于侧向冲击，MPS 在 0.20 到 0.48 之间（左侧半球）和 0.22 到 0.51 之间（右侧半球）。对于前向冲击，区域累积应变损伤量 (CSDM20) 在 0.01 到 0.87 之间。对于侧向冲击，CSDM20 在 0.36 到 0.99 之间（左侧半球）和 0.09 到 0.93 之间（右侧半球）。
鉴于神经功能与解剖结构有关，并且大多数基于模型的损伤指标都集中在整个大脑的反应上，本研究开发了一个解剖学上准确的人脑模型来捕捉区域反应。模型验证与现有模型相当。该模型提供了足够的解剖细节，可以描述不同冲击条件下的脑区反应。
