Methodological advances for the quantification of body movement and concurrently measured brain activity of exergame players

Abstract

The aging population and a projected decline in the number of healthcare personnel necessitate the development and increased use of health technology to halt or delay functional decline in older adults connected to aging. Exergames in particular show great promise in both exercise and clinical contexts to train and improve both physical and cognitive function. However, there has been little attention to the specifics of movements and brain activity elicited during exergaming, how these are influenced by exergame settings and not in the least, how to concurrently measure and process movement and brain activity. The aim of this dissertation is to contribute to the development of exergames as a tool for training and rehabilitation by investigating the effect of game settings on movement characteristics and cortical activity, and of movement task on automated EEG cleaning performance. Paper I investigated how two key game elements, game speed and the presence of obstacles, influence movement characteristics in 15 older adults playing a step-based balance training exergame. The task consisted of moving sideways to catch falling grapes and to avoid obstacles (falling branches), if present. Occasionally appearing chickens were caught by raising at least one arm over the head. Participants played the game for eight 2 min trials in total, at two speed settings and with or without obstacles. The 3D position of 22 retroreflective markers fixed to anatomical landmarks was captured using a motion capturing system. Calculated movement characteristics included step size, step frequency, single leg support, arm lift frequency, and horizontal trunk displacement. An increase in game speed resulted in a decrease in mean single support time, step size, and arm lift frequency, and an increase in cadence, game score, and number of error messages. The presence of obstacles resulted in a decrease in single support ratio, step size, cadence, frequency of arm lifts, and game score. Furthermore, an increase in step size from the first to the second trial repetition was observed. These results show that both game speed and the presence of obstacles altered movement characteristics with some changes considered beneficial and others detrimental for the effectiveness of balance training. Paper II aimed to assess whether concurrent electrophysiological measurements during exergaming are feasible and if so, whether cortical activity changes with additional cognitive elements. Twenty-four young adults first performed self-paced sideways leaning movements, directly followed by two blocks of exergames in which the same movement as input was used. The task of the exergame was to complete a 5 by 5 puzzle matrix. Puzzle pieces were selected by leaning towards them. The exergames were played in two difficulty levels. At the easy level, only the correct piece was shown, while two pieces were presented for the more difficult level. Brain activity was recorded using a 64-channel passive EEG system. Results showed that it is feasible to record brain activity in young adults while playing exergames. Five spatially different clusters of independent components were identified located frontal, bilateral central, and bilateral parietal. Significantly higher absolute theta power in the more difficult exergaming condition was found compared to the easy level and the self-paced movement. Both central clusters showed a significant increase in absolute alpha-2 power in the exergaming conditions compared to the self-paced movements. In Paper III, the effect of task on the EEG artifact removal abilities of artifact subspace reconstruction was assessed. Using state-of-the art preprocessing algorithms is a precondition for artifact contaminated EEG recorded during more movement intensive exergames. However, the effect of the task on artifact subspace reconstruction has not yet been assessed. EEG recorded during three tasks was preprocessed manually and by the use of artifact subspace reconstruction using 10 cut-off parameters, which can determine the rigor of the artifact reconstruction. The mean cut-off parameter equivalent to the ratio of EEG removed in manual cleaning was strictest for the walking task. Quality indexes of independent components which give information about the repeatability of independent component decompositions were best for the walking and worst for the single-leg stance task across all cut-off parameters. Furthermore, quality indexes of independent components reached a maximum plateau for cut-off parameters of 10 and higher. Dipolarity was largely unaffected by the choice of the cut-off parameter. The number of independent components within each task remained constant, regardless of the cut-off parameter used. Surprisingly, ASR performed better in motor tasks compared to non-movement tasks. Furthermore, there was no benefit of using cut-off parameters less than 10. The combined results of these papers show that game settings in exergames influence both brain activity and movement characteristics of the players, and this can be either beneficial or detrimental to the desired training effects. Therefore, an informed approach is needed in order to achieve the intended benefit and effectiveness when choosing specific game settings in an exergame. Furthermore, simultaneous measurement of cortical activity while exergaming is possible, despite the presence of movement artifacts. The cleaning performance of artifact subspace reconstruction was best in the most movement intensive task, indicating its applicability for use in EEG collected during exergaming. This result paves the way for continued development of state-of the-art EEG preprocessing algorithms that enable analysis of cortical activity also in movement intensive exergames, thereby allowing concurrent analysis of brain activity in specific areas during playing. The resulting knowledge from all three papers can contribute to the continued development of more targeted and effective exergames and enable future studies to investigate movement characteristics and brain activity concurrently.