Machine Learning Enabled Wearable Brain Deformation Sensing System

Brain deformation - the primary cause of traumatic brain injury (TBI) - occurs during fall, automobile accident, brain surgery, or explosion (i.e., pressurized airflow) [1] . Mechanical impact causes strain energy that leads to tissue displacement. Researchers have attempted to characterize the brain deformation for diagnosis and prevention of concussion-related TBI [2] . It is especially important to measure microscale deformation because even a few tens of micrometer brain deformation may have direct neuropsychiatric and neuro-degenerative consequences [3] - [6] . Another effort to minimize brain deformation can be found in intracranial surgeries. The deformation is inevitable but can be minimized by designing a better apparatus and using advance stereotactic techniques [7] - [9] . As such, there are a few methods to measure brain deformation today [8] , [10] - [12] . Computational models and imaging technologies (e.g., FEM (finite element method) modeling, magnetic resonance imaging (MRI) are such examples. However, because the brain is viscoelastic [13] , these technologies lack 1) detailed information regarding micro-scale brain deformation and 2) real-time measurement capability. Our objective is to develop a novel sensing system capable of measuring microscale brain deformation in real-time. In this abstract, we report a wireless brain deformation sensing system based on the magnetic tunnel junction sensor and an implantable soft magnet combined with machine learning.