Table 2. Center of Mass (COM) Assessment Protocol and Results.

Study	Gait COM Protocol	Assessment Time Points	Study Results
Fino, et al6 (2016)	2 six-axis IMUs aligned in the mediolateral, vertical, and anterior posterior directions with data sampled at 128 Hz during single task and dual task gait. 18m walkway	Assessed weekly for six weeks and a one-year follow-up assessment.	Single Task: no differences in stability or variability between groups; no significant main effects of group, week or task found for stride time, variability, λs-Trunk, or λs-Head. Dual Task: gait speed was slower than single task gait speed, with increased speed over time.
Parker, et al10 (2008)	External markers and estimated joint centers were used to calculate 3-dimensional motion for individual body segments and locations of segmental COM. Two COM variables were examined: (1) the COM displacement in the medial-lateral direction and (2) the maximum separation between COM and COP of the supporting foot in the anterior direction. The relationship between the whole-body COM and the base of support (shown to be a sensitive measure of gait imbalance), 10m walkway	Assessed 48 hours after injury concussed), day 2 (non-concussed), day 5 (all), day 14 (all), and day 28 (all)	Gait imbalance during the divided attention condition was marked by greater sway and sway velocity of the whole-body COM that was maintained for up to 28 days following injury.
Catena, et al2 (2009)	29 retroreflective markers attached to anatomical landmarks while 3D marker trajectories were taken with eight camera motion tracking system at 60Hz, then filtered with low-pass fourth order Butterworth filter at cutoff frequency of 8 Hz. Marker position data was used to locate segmental COM of a thirteen-link model: head, trunk, two upper arms, two lower legs, pelvis, two thighs, two shanks, two feet.	Assessed 48 hours, on the 6th day, 14th day, and 28th day post-injury.	Concussed individuals significantly reduced peak anteroposterior velocity during dual task walking on day 2. Peak mediolateral velocity was significantly reduced by day 14 during short obstacle crossing.
Howell, et al7 (2015)	29 retroreflective markers placed on bony landmarks of the patient with whole body motion analysis performed using a 10-camera motion analysis system at a sampling rate of 60 Hz capturing and reconstructing 3D trajectory of each marker. Marker trajectory data was low-pass filtered using the fourth-order Butterworth filter with cutoff frequency set to 8 Hz. whole body COM positions were calculated as the weighted sum of all 13 body segments to represent the whole body. 15m walkway	Assessed within 72 hours of injury and 1 week, 2 weeks, 1 month, and 2 months post-injury.	Concussion group: Significant differences were found in group-time interaction between dual-task walking for mediolateral displacement of COM and COM medial-lateral velocity. Significant worsening of COM control after return to activity was also illustrated during dual-task walking. Overall mean return to activity mediolateral displacement was significantly greater than controls for same time point measurements when dual-tasking gait. The percent change value of medial-lateral velocity during dual-task walking was significantly greater. Peak COM anterior velocity was also decreased in percent change value between pre- and post-return to activity while dual-tasking gait There was a significant group-time interaction pre- and post-return to activity in clinical symptom scores. Pre-Return timepoints between tests 2 and 1 changed significantly more than that of controls which showed little to no change for either testing interval. Additionally, mean pre- and post-return to activity changes were significantly different than controls for clinical symptoms.

Abbreviations: IMU: inertial measurement unit, COP: Center of Pressure, λs: Lyapunov exponents