EFFECT OF FILTERING KINEMATICS ON FINITE ELEMENT SIMULATIONS OF HEAD IMPACTS IN HIGH SCHOOL FEMALE LACROSSE

The recent Consensus Head Acceleration Measurement Practices (CHAMP) Conference recommends the reporting of filter characteristics for head impact sensors, but CHAMP did not prescribe a specific filter and/or cut-off frequency as selection depends on device and application. A previous study reported 14 video-verified impacts to the face/jaw region of 8 high school female lacrosse players recorded by custom-fitted Stanford Instrumented Mouthguard (MiG) sensors. The raw kinematics data were originally filtered at 160 Hz using a 4th order Butterworth filter as specified by Stanford. Separately, the current study filtered the raw kinematics data at previously reported cut-off frequencies of 50, 100 and 200 Hz, and compared peak values. In addition, the filtered kinematics were used to simulate the impacts using a finite element (FE) human head model and 95^th percentile stresses and strains within the brain were compared. Lowering the cut-off frequency of the low-pass filter substantially reduced peak linear and angular accelerations, whereas peak angular velocity was less affected. In addition, a flow-on effect was observed as lowering the filtering cut-off frequency reduced 95th percentile stresses and strains within the brain. While low-pass filtering is a common approach to remove high-frequency noise from kinematics signals, information regarding the actual signal may be lost from over-filtering. Future studies using instrumented mouthguard data to investigate acceleration-based injury metrics, or drive FE human head models, should carefully consider filtering methods.