Head Impact Sensor Triggering Bias Introduced by Linear Acceleration Thresholding

• Published in: Annals of biomedical engineering Vol. 49; no. 12; pp. 3189 - 3199
• Main Authors: Wang, Timothy, Kenny, Rebecca, Wu, Lyndia C.
• Format: Journal Article
• Language: English
• Published: Cham Springer International Publishing 01.12.2021; Springer Nature B.V
• Subjects: Acceleration; Accelerometry - instrumentation; Adolescent; Adult; Bias; Biochemistry; Biological and Medical Physics; Biomechanical Phenomena; Biomedical and Life Sciences; Biomedical Engineering and Bioengineering; Biomedicine; Biophysics; Classical Mechanics; Contact sports; Exposure; Female; Football; Head; Head Movements; Humans; Kinematics; Male; Mouth Protectors; Original Article; Protective equipment; Recording; Rotation; Sensors; Soccer; Head impact exposure; Triggering bias; Subconcussive impacts; Linear acceleration threshold; Head impact sensors; Impact triggering
• ISSN: 0090-6964; 1573-9686
• DOI: 10.1007/s10439-021-02868-y

Contact sports players frequently sustain head impacts, most of which are mild impacts exhibiting 10–30 g peak head center-of-gravity (CG) linear acceleration. Wearable head impact sensors are commonly used to measure exposure and typically detect impacts using a linear acceleration threshold. However, linear acceleration across the head can substantially vary during 6-degree-of-freedom motion, leading to triggering biases that depend on sensor location and impact condition. We conducted an analytical investigation with impact characteristics extracted from on-field American football and soccer data. We assumed typical mouthguard sensor locations and evaluated whether simulated multi-directional impacts would trigger recording based on per-axis or resultant acceleration thresholding. Across 1387 impact directions, a 10g peak CG linear acceleration impact would trigger at only 24.7% and 31.8% of directions based on a 10 g per-axis and resultant acceleration threshold, respectively. Anterior impact locations had lower trigger rates and even a 30 g impact would not trigger recording in some directions. Such triggering biases also varied by sensor location and linear-rotational head kinematics coupling. Our results show that linear acceleration-based impact triggering could lead to considerable bias in head impact exposure measurements. We propose a set of recommendations to consider for sensor manufacturers and researchers to mitigate this potential exposure measurement bias.