A leading cause of traumatic brain injury (TBI) is intracranial brain deformation due to mechanical impact. This deformation is viscoelastic and differs from a traditional rigid transformation. In this paper, we describe a machine learning-enabled wireless sensing system that predicts the trajectory of intracranial brain deformation. The sensing system consists of an implantable soft magnet and an external magnetic sensor array. It has a three-dimensional sensing volume of 12 ´ 12 ´ 4 mm3, with a spatial resolution of 1 µm and a temporal resolution of 5 µs. The machine learning algorithm predicts the brain deformation by interpreting the magnetic sensor outputs created by the change in position of the implanted soft magnet. Three different machine learning models were trained on calibration data: (1) random forests, (2) k-nearest neighbors, and (3) a multi-layer perceptron-based neural network. These models were validated using a needle insertion into PVC gels, as well as blast exposure to live and dead rat heads. The in vitro gel deformation predicted by the machine learning models showed excellent agreement with the reference camera measurements; absolute error was 138 µm, Fréchet distance was 372 µm, and the normalized Procrustes disparity was 0.034. The in vivo brain deformation prediction also showed promising results compared to the theoretical model; average absolute error was 86 µm, average Fréchet distance was 129 µm, and average Procrustes disparity was 0.127. The overall accuracy of the machine learning-enabled sensing system showed over 92%. These results suggest that the machine learning-enabled wireless intracranial brain deformation sensing system can be an effective tool for studying TBI due to in situ and real-time brain deformation prediction.
Sign-in or become an IEEE member to discover the full contents of the paper.