The Impact of Drop Test Conditions on Brain Strain Location and Severity: A Novel Approach Using a Deep Learning...

Area of Science:

• Biomechanics
• Neuroscience
• Machine Learning

Background:

• Contact sports like rugby carry a high risk of traumatic brain injuries (TBI) due to severe head impacts.
• Previous research confirms a strong correlation between high-impact head accelerations and the brain strains leading to concussions.

Purpose of the Study:

• To investigate the influence of various laboratory-controlled drop test parameters on regional maximum principal strain (MPS) predictions in the brain.
• To utilize a trained convolutional neural network (CNN) for predicting brain strains under different impact conditions.

Main Methods:

• A novel approach employing a trained convolutional neural network (CNN) to predict regional peak and mean maximum principal strain (MPS).
• Analysis of laboratory-controlled drop test parameters and their effect on brain strain predictions.
• Comparison of MPS values resulting from forehead versus side head impacts, and the impact of including a neck structure.

Main Results:

• Impact location significantly affects MPS; side impacts generate higher regional MPS than forehead impacts.
• Forehead impacts at 0° and 45° surface angles yielded the lowest region-averaged MPS.
• The absence of a neck in drop tests led to reduced regional peak and mean MPS values.
• A complex relationship exists between drop test parameters and predicted regional brain strains.

Conclusions:

• Deep learning models, specifically CNNs, offer a powerful tool for detailed analysis of how drop test conditions influence regional brain strain (MPS).
• The findings provide crucial insights for developing advanced protective headgear to mitigate brain strain from head accelerations in sports.