Stanford researchers at the Camarillo Lab have developed a neural-network based model that can provide real-time calculation of brain strain based on instrumented mouthguard kinematics signals. This fast and accurate calculation can help better protect athletes wearing the instrumented mouthguard disclosed in Stanford docket S15-432.
Current approaches for head impact measurement such as finite element analysis (FEA) are much slower and do not use real-time models based on the complex dynamics of the head-brain interface to predict and visualize brain strain. This new neural-network model is dramatically faster, easier to interpret, and can calculate detailed spatial resolution of brain strain, showing metrics (maximum principal strain) in each of the elements of the brain.
Flowchart introduction of the model.
The visualization of the effectiveness and accuracy of the deep learning head model.
Stage of Development
- Mild traumatic brain injury (mTBI) diagnostics
- Can be used by Sports teams and Researchers
- Accurate and real-time calculation of brain strain calculation
- Fast calculation - the new model is much faster (
- Detailed spatial resolution of brain strain, showing metrics (maximum principal strain strain) in each of the elements of the brain
- Easy to interpret – users without FEA experience can easily understand results
- Zhan, Xianghao, Yuzhe Liu, Samuel J. Raymond, Hossein V. Alizadeh, August G. Domel, Olivier Gevaert, Michael Zeineh, Gerald Grant, and David B. Camarillo. Deep Learning Head Model for Real-time Estimation of Entire Brain Deformation in Concussion. arXiv preprint arXiv:2010.08527 (2020)
Instrumented Mouthguard to Determine Accurate Head Motion During ImpactsS15-432
Instrumented Mouthguard to Determine Accurate Head Motion During Impacts
Ultrathin Flexible Multilead Electrocardiogram Bandage for Arrythmia Diagnosis and MonitoringS19-061
Ultrathin Flexible Multilead Electrocardiogram Bandage for Arrythmia Diagnosis and Monitoring
A Novel Approach for Detecting Head Collisions in SportsS13-015
A Novel Approach for Detecting Head Collisions in Sports