RESUMEN
PURPOSE: This study aimed to quantify differences between head acceleration measured by a helmet-based accelerometer system for ice hockey and an anthropometric test device (ATD) to validate the system's use in measuring on-ice head impacts. METHODS: A Hybrid III 50th percentile male ATD head and neck was fit with a helmet instrumented with the Head Impact Telemetry (HIT) System for hockey and impacted at various speeds and directions with different interfaces between the head and helmet. Error between the helmet-based and reference peak accelerations was quantified, and the influence of impact direction and helmet-head interface was evaluated. Regression equations were used to reduce error. System-reported impact direction was validated. RESULTS: Nineteen percent of impacts were removed from the data set by the HIT System processing algorithm and were not eligible for analysis. Errors in peak acceleration between the system and ATD varied from 18% to 31% and from 35% to 64% for linear and rotational acceleration, respectively, but were reduced via regression equations. The relationship between HIT System and reference acceleration varied by direction (P < 0.001) and head-helmet interface (P = 0.005). Errors in impact azimuth were approximately 4%, 10%, and 31% for side, back, and oblique back impacts, respectively. CONCLUSIONS: This is the first comprehensive evaluation of peak head acceleration measured by the HIT System for hockey. The HIT System processing algorithm removed 19% of the impacts from the data set, the correlation between HIT System and reference peak resultant acceleration was strong and varied by head surface and impact direction, and the system error was larger than reported for the 6-degree-of-freedom HIT System for football but could be reduced via calibration factors. These findings must be considered when interpreting on-ice data.