TY - JOUR T1 - Video Assessment of Head Impact Exposure in American Football JF - bioRxiv DO - 10.1101/235432 SP - 235432 AU - Calvin Kuo AU - Lyndia Wu AU - Jesus Loza AU - Daniel Senif AU - Scott C. Anderson AU - David B. Camarillo Y1 - 2017/01/01 UR - http://biorxiv.org/content/early/2017/12/16/235432.abstract N2 - Previous research has sought to quantify head impact exposure using wearable kinematic sensors. However, validation studies have shown that most sensors suffer from poor accuracy in estimating impact severity and count, motivating the need for additional independent impact exposure quantification. Here, we implemented a tiered video assessment to collect American football head impact exposure data. We also developed a new exposure metric unique to American football quantifying number of impacts per player-play, and found significantly higher exposure during games (0.40, 95% CI: 0.33-0.47) than practices (0.26, 95% CI: 0.23-0.30) (p<0.05). Using the more traditional impacts per player-hour metric, we observed higher exposure during practices (4.7) than games (3.6) due to the greater number of player-plays per hour (17.8 practices vs. 9.1 games). Thus, our exposure metric accounts for variability in on-field participation, providing more accurate depictions of impact likelihood when in play. We further coupled our video assessment with an instrumented mouthguard, and found median impact linear acceleration angular velocity, and angular acceleration magnitudes of 19.0g, 9.9rad/s, and 1358.8rad/s2 respectively during practices, and more severe 23.7g, 13.4rad/s, and 1483.5rad/s2 respectively during games (p<0.5 in linear acceleration magnitude). Our results agree with previous findings showing higher exposure and more severe impacts during games. However, our impact directionality results differed substantially, with 1.5% identified as rear impacts in video assessment, and 30.9% identified as rear impacts in previous sensor-based exposure studies. This discrepancy demonstrates the need for collecting independent, unbiased impact exposure data to compare against sensor-based findings. ER -