Ustinova KI, Ingersoll CD, Cassavaugh ND Development of 3D videogame to train the whole body coordination in patients with TBI Thanks for your interest to this topic. I’ll try not to be boring while talking about development of 3D videogame for patients with Tbi and effect of practicing with this game on the whole body coordination. Ustinova KI, Ingersoll CD, Cassavaugh ND

BACKGROUND & PURPOSE TBI affects the central and executive mechanisms underlying arm and postural coordination. Little attention is paid to its restoration by conventional rehabilitation As alternative, a customized 3D game was developed to improve coordination in TBI patients Effect of a short-term gaming practice on the arm and postural coordination of patients with TBI was analyzed First, what was a purpose of development of such game? We all know, that TBI disrupts the central and executive mechanisms underlying the coordination of the entire body, and specifically coordination of arm and postural segments, trunk and legs. It results in postural instability when performing arm movements, increases the risk of falling, deteriorates motor skills, and eventually decreases the quality of life of TBI survivors [2,3,4]. Despite the importance of this ability, You will be surprised how little attention is paid to its restoration by conventional rehabilitation, which generally treats the affected upper extremities, postural control, and gait separately. This is partially because of the complexity of the whole-body movements, and sometimes difficulties addressing a specific problem with real world coordination tasks, and adapting these tasks to the abilities and needs of TBI survivors. To address this issue, we developed customized virtual reality (VR)-based gaming exercise, game. Vr game is very convenient tool, since poses no risk to participant, allows acknowledge his/her success according to the abilities, and at the same time motivating and fun, compared to regular exercises. We tested the effect of short-term gaming practice on arm-postural coordination in patients with TBI.

Thirteen patients with mild-to-moderate manifestations of TBI PARTICIPANTS Thirteen patients with mild-to-moderate manifestations of TBI Age 32±6.7 (6 males and 7 females) Ataxia Score according to Klockgether 5-12pts (35 severe ataxia) Berg Balance test 39-55pts (45 – high fall risk) Functional Gait Assessment Test 12-29pts (22 – high fall risk) No severe visual perception deficits No severe cognitive deficits 13 patients with mild-to-moderate ataxia participated in the study, they were 32-year of age on average, and a subject pool included 6 males and 7 females. All participants had mild-to-moderate ataxia with scores ranging from 5 to 12, where the score 35 identifies severe ataxia. They had impaired balance with scores from 39 to 55 on the berg balance test, with 45 identifying high risk of falling. Gait impairments were scored from 12-29 with 22 corresponding the high risk of fall again. Participants had normal or corrected vision, and no severe cognitive deficits which interfered with their ability to understand instructions and perform the task.

GAME DESIGN The game Octopus The participants played the game, designed to challenge postural stability while reaching to intercept a moving target. We called the game Octopus. All actions occur in an underwater world populated with seaweeds and corals (Fig. 1A-B). The main character, Octopus, is located in the middle of the screen. Octopus blows bubbles towards the participant, whose presence in the underwater landscape is indicated by the right and left hand avatars. The avatars were created with 3 markers attached to each hand, captured by the system for motion analysis, and then synchronized with the VR gaming scenario. The image was projected in 3D format onto an 82-inch screen and was viewed by the participant via shutter glasses. The gaming task was to reach and pop (intercept) as many bubbles (targets) as possible with the left or right hand. The Octopus blows bubbles regularly every 4 s at a speed of 1.5 m/s. Each successful bubble trajectory interception is rewarded with points, which accumulate throughout the gaming trial and serve as the criteria of performance success (performance score). Each next trial participant needs to exceed the score 10% to have a new character appears on the Octopus landscape. It may be a fish, diver, treasure chest and so on. The ultimate goal is to collect as many characters as possible and get the final reward, an animated Mermaid. Gaming trials last 90 sec with 25 reach-to-pop movements to be performed. Practice consisted of 10 trials, with a 30-min interval, followed by 2 retention trials.   Task – to pop the bubbles, with the left of right hand avatar Bubble speed 1.5 m/s, reward – appearance of a new character Gaming trial - 90 s, ~20-25 reach-to-pop movements Practice consisted of 10 trials, with a 30-min interval, followed by 2 retention trials

GAME DESIGN The bubble trajectories 1 2 3 4 5 20-25 cm 15-18 cm Once launched, bubbles randomly followed 1 of 5 radial (circular) trajectories (Fig. 2A-B), designed so that the target in overhead position corresponded to participants height with arm raised up. At the shoulder level in the frontal plane, the target is 15-18 cm beyond the length of the arm, outstretched to the right or left side. In the sagittal plane, the target at the shoulder level is 25-30 cm beyond the length of the arm, outstretched forward. These reaching distances are considered as the lower borders for norms on the Berg Balance test on the Multi-Directional Reach Test. The tests which were design to test body stability during performance of functional arm movements. Thus, while playing the game, a participant had a few options. He might pop the bubble in overhead position, that required minimum postural displacement, but took longer, or he might intercept  the bubble early, but for that he needs to lean forwad, thereby perturbing stability. For the purpose of this study, the reach-to-pop bubble #3 , which is in strictly sagittal plane was nalyzed. Reaching-to-pop bubble #3 with the dominant hand was analysed

DATA COLLECTION & ANALYSIS Kinematic analysis included: Arm movement time (s) Arm trajectory curvature (arm trajectory length/shortest path to the target) Arm-postural coordination (Principal Component Analysis, PCA) Angular displacements of 9 body segments (i.e., 2 hands, 2 forearms, 2 upper arms, trunk, and 2 legs) were included PCA percent variance, (PC I mainly) PC loading of each segment (significant if >0.7) We recorded kinematics of the whole body, and from that analyzed arm movement time – how fast arm moved to pop a bubble; arm trajectory curvature; this parameter tells how much the hand was deviated from the shortest path to the target, and is indirect measurement of arm precision. And the we analyzed arm-postural coordination with the principal component analysis. The analysis included angular displacement of 9 body segments, which could potentially be involved in gaming task performance. Those are both hands, forearms, upper arms, trunk, and both legs. The leg segment included both thigh and shank. Analysis provided two major parameters: Principal Component percent variance, and Principal Component loadings. I will clarify their means later, when we get to the discussion of results.

RESULTS Trajectories of the arm (hand), trunk (C7), and legs (hip) Trial 1 Trial 10 While practicing the game, the participants were not instructed on how to move to catch a bubble successfully. The bubble trajectory could be intercepted using different combinations of arm and postural segment displacements. These figure illustrate the trajectories of major body segments involved in task performance from one representative participant. The grey body serves as a link between, to illustrate participant’s movements. His initial attempt to reach the bubble was characterized by a longer and less-accurate hand movement (Fig. 3A). The target trajectory was intercepted at an almost overhead position that required minimal postural involvement. By the end of the practice session on the ninth trial (Fig. 3B), the hand trajectory became shorter, less curved, and the bubble trajectory was intercepted earlier than on the first trial. To reach the bubble, the participant leaned forward and used his leg to counterbalance the forward body shift. This later strategy revealed the greater involvement of postural segments into arm transport. Such tendency was observed in most of our patients.

RESULTS As a result of short-term practice: Gaming task performance improved Arm movement time decreased Trajectory curvature decreased As a result most of them improved performance score, arm movement time and trajectory curvature. That is seen from the figures below. Performance score, which was a number of points earned by participants per trial improved significantly by the 6th trial, with retention over a 30-min break. The happened to movement time, which was reduced upon completion of 2/3 of gaming trials. Again, participants demonstrated similar results after break. Finally, the movements became more accurate and deviated less from the shortest path, causing reduction in trajectory curvature. Overall, participants improved movement performance on all these parameters, mostly after completing at least half of trials. * difference between the first trial and the trial where significant changes occurred; # difference between the first and the last trial;

RESULTS As a result of short-term practice: Inter-segmental coordination improved (PC1) The movement coordination and relative contribution of different body segments to arm transport were analysed using PCA. About 90% of the variance in the angular displacements of the 9 segments was accounted for by the first 3 PCs (Fig. 5A). Her we were looking for amount of variance explained by the first PC. Because the PC 1 explain coordination between segments during performance of the movement itself. The amount of variance explained by PC1 was significantly different for reaches-to-pop during the first gaming trial (39% ± 8%) than for those during the last trial (68% ± 9%). The percentage of angular variance associated with PC1 provided an evidence of coupling between the segments when popping the bubbles. This coupling was increased by the end of the practice session (F1,10=1.95, p=0.023). The first significant change in percentage was observed by the eight trial (p=0.001) with a tendency to retain acquired coordination over retention interval. Thus we can say that movements in our participants became more coordianted by the end of gaming practice. * difference between the first trial and the trial where significant changes occurred; # difference between the first and the last trial;

RESULTS As a result of short-term practice: Contribution (PC loadings) of postural segments into task performance increased The next thing we analyzed was a contribution of each of participating segments into task performance. Here only 5 out of 9 segments are presented as having loading coefficient 0.7 and above. These segments include: forearm and upper arm from the dominant side, trunk and both dominant and non-dominant legs. Again we are interested in the loadings within the first PC only (black bars). Asterisks identify significant difference between the first and last trials. If you look at postural segments (trunk and legs) you can see that the pattern of their loading coefficients was markedly different as participants progressed from th e beginning to the end of the gaming session. These coefficients were much smaller on PC1 during the first trial, indicating their weak coupling to arm motion. High loadings for the legs and trunk in PC1 on the last trial, combined with the fact that PC1 explained 68% of the variance, indicated that these segments were tightly synchronized with arm motion, forming the synergetic pattern of the whole-body motion. * difference between the first and the last trial

DISCUSSION & CONCLUSION Upon completion ¾ of gaming tasks participants: Improved game performance, arm movement time and precision Improved movement coordination Increased contribution of postural segments into movement performance Improved on clinical tests: forward reach (10/13 participants), and single-leg stance time (9/13 participants) Changes were retained over the 30-min retention interval These results support the feasibility of using the custom-made 3D game for retraining of arm-postural coordination disrupted by TBI. In conclusion:     

ACKNOWLEDGEMENTS Supported by a grant from the US Department of Defense Thank you to Chris Hausbeck, Jessica Gardon-Rose, Amanda Schafer