Time Window of Head Impact Kinematics Measurement for Calculation of Brain Strain and Strain Rate in American Football

Liu, Yuzhe; Domel, August G.; Cecchi, Nicholas J.; Rice, Eli; Callan, Ashlyn A.; Raymond, Samuel; Zhou, Zhou; Zhan, Xianghao; Zeineh, Michael; Grant, Gerald A.; Camarillo, David B.

doi:10.48550/arxiv.2102.05728

Cited by 4 publications

(3 citation statements)

References 55 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A 220 ms window was used based on two studies of the dynamic characteristics of brain strain during impact. 24,37 Both studies found that long-duration (above 100 ms) data recording was needed to simulate an appropriate brain strain response which included the peak strain values. Resultant kinematic traces were subsequently normalized based on the peak value, following which the mean kinematic profile shape was found.…”

Section: Data Acquisition and Post Processingmentioning

confidence: 99%

Drop Test Kinematics Using Varied Impact Surfaces and Head/Neck Configurations for Rugby Headgear Testing

et al. 2022

View full text Add to dashboard Cite

World Rugby employs a specific drop test method to evaluate headgear performance, but almost all researchers use a different variation of this method. The aim of this study was, therefore, to quantify the differences between variations of the drop testing method using a Hybrid III headform and neck in the following impact setups: (1) headform only, with a flat steel impact surface, approximating the World Rugby method, (2 and 3) headform with and without a neck, respectively, onto a flat MEP pad impact surface, and (4) headform and neck, dropped onto an angled MEP pad impact surface. Each variation was subject to drop heights of 75–600 mm across three orientations (forehead, side, and rear boss). Comparisons were limited to the linear and rotational acceleration and rotational velocity for simplicity. Substantial differences in kinematic profile shape manifested between all drop test variations. Peak accelerations varied highly between variations, but the peak rotational velocities did not. Drop test variation also significantly changed the ratios of the peak kinematics to each other. This information can be compared to kinematic data from field head impacts and could inform more realistic impact testing methods for assessing headgear.

show abstract

Section: Data Acquisition and Post Processingmentioning

confidence: 99%

Drop Test Kinematics Using Varied Impact Surfaces and Head/Neck Configurations for Rugby Headgear Testing

et al. 2022

View full text Add to dashboard Cite

show abstract

“…In this study, to cover a broader range of head impacts, we included 3,161 impact data from four different impact sources: 1) 2,130 football-like impacts simulated by a finite element model of a Hybrid III anthropomorphic test device headform (ATD) (labeled as HM for head model) [26], [27]; 2) 302 college football impacts measured by the Stanford Instrumented Mouthguard (labeled as CF) [28]- [30]; 3) 457 mixed martial arts impacts measured by the Stanford Instrumented Mouthguard (labeled as MMA) [7], [31]; 4) 272 reconstructed impacts from the National Association for Stock Car Auto Racing (labeled as NASCAR) [32], [33].…”

Section: A Datasets: Head Kinematics and Brain Finite Element Modelmentioning

confidence: 99%

Data-driven decomposition of brain dynamics with principal component analysis in different types of head impacts

Zhan¹,

Liu

Cecchi³

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

Strain and strain rate are effective traumatic brain injury predictors. Kinematics-based models estimating these metrics suffer from significant different distributions of both kinematics and the injury metrics across head impact types. To address this, previous studies focus on the kinematics but not the injury metrics. We have previously shown the kinematic features vary largely across head impact types, resulting in different patterns of brain deformation. This study analyzes the spatial distribution of brain deformation and applies principal component analysis (PCA) to extract the representative patterns of injury metrics (maximum principal strain (MPS), MPS rate (MPSR) and MPS×MPSR) in four impact types (simulation, football, mixed martial arts and car crashes). Methods: We apply PCA to decompose the patterns of the injury metrics for all impacts in each impact type, and investigate the distributions among brain regions using the first principal component (PC1). Furthermore, we developed a deep learning head model (DLHM) to predict PC1 and then inverse-transform to predict for all brain elements. Results: PC1 explained > 80% variance on the datasets. Based on PC1 coefficients, the corpus callosum and midbrain exhibit high variance on all datasets. We found MPS×MPSR the most sensitive metric on which the top 5% of severe impacts further deviates from the mean and there is a higher variance among the severe impacts. Finally, the DLHM reached mean absolute errors of < 0.018 for MPS, < 3.7s −1 for MPSR and < 1.1s −1 for MPS×MPSR, much smaller than the injury thresholds. Conclusion: The brain injury metric in a dataset can be decomposed into mean components and PC1 with high explained variance. Significance: The brain dynamics decomposition enables better interpretation of the patterns in brain injury metrics and the sensitivity of brain injury metrics across impact types. The decomposition also reduces the dimensionality of DLHM.

show abstract

“…Only within the bounds of the data used to train the model can a network be relied upon to provide any useful information. Even within these bounds though, there has already been a number of useful applications of deep learning in the physical sciences [7,8,9,10,11,12,13,14] and this manner of application will continue to develop as more data is generated for scientific purposes. Relying on a data-driven model to provide mechanistic predictions though, creates a number of problems.…”

Section: Introductionmentioning

confidence: 99%

Applying physics-based loss functions to neural networks for improved generalizability in mechanics problems

Raymond¹,

Camarillo²

2021

Preprint

Self Cite

View full text Add to dashboard Cite

Physics-Informed Machine Learning (PIML) has gained momentum in the last 5 years with scientists and researchers aiming to utilize the benefits afforded by advances in machine learning, particularly in deep learning. With large scientific data sets with rich spatio-temporal data and high-performance computing providing large amounts of data to be inferred and interpreted, the task of PIML is to ensure that these predictions, categorizations, and inferences are enforced by, and conform to the limits imposed by physical laws. In this work a new approach to utilizing PIML is discussed that deals with the use of physics-based loss functions. While typical usage of physical equations in the loss function requires complex layers of derivatives and other functions to ensure that the known governing equation is satisfied, here we show that a similar level of enforcement can be found by implementing more simpler loss functions on specific kinds of output data. The generalizability that this approach affords is shown using examples of simple mechanical models that can be thought of as sufficiently simplified surrogate models for a wide class of problems.

show abstract

Time Window of Head Impact Kinematics Measurement for Calculation of Brain Strain and Strain Rate in American Football

Cited by 4 publications

References 55 publications

Drop Test Kinematics Using Varied Impact Surfaces and Head/Neck Configurations for Rugby Headgear Testing

Drop Test Kinematics Using Varied Impact Surfaces and Head/Neck Configurations for Rugby Headgear Testing

Data-driven decomposition of brain dynamics with principal component analysis in different types of head impacts

Applying physics-based loss functions to neural networks for improved generalizability in mechanics problems

Contact Info

Product

Resources

About