Music meets robotics: a prospective randomized study on motivation during robot aided therapy

Abstract

Background: Robots have been successfully applied in motor training during neurorehabilitation. As music is known to improve motor function and motivation in neurorehabilitation training, we aimed at integrating music creation into robotic-assisted motor therapy. We developed a virtual game-like environment with music for the arm therapy robot ARMin, containing four different motion training conditions: a condition promoting creativity (C+) and one not promoting creativity (C-), each in a condition with (V+) and without (V-) a visual display (i.e., a monitor). The visual display was presenting the game workspace but not contributing to the creative process itself. In all four conditions the therapy robot haptically displayed the game workspace. Our aim was to asses the effects of creativity and visual display on motivation.

Methods: In a prospective randomized single-center study, healthy participants were randomly assigned to play two of the four training conditions, either with (V+) or without visual display (V-). In the third round, the participants played a repetition of the preferred condition of the two first rounds, this time with a new V condition (i.e., with or without visual display). For each of the three rounds, motivation was measured with the Intrinsic Motivation Inventory (IMI) in the subscales interest/enjoyment, perceived choice, value/usefulness, and man-machine-relation. We recorded the actual training time, the time of free movement, and the velocity profile and administered a questionnaire to measure perceived training time and perceived effort. All measures were analysed using linear mixed models. Furthermore, we asked if the participants would like to receive the created music piece.

Results: Sixteen healthy subjects (ten males, six females, mean age: 27.2 years, standard deviation: 4.1 years) with no known motor or cognitive deficit participated. Promotion of creativity (i.e., C+ instead of C-) significantly increased the IMI-item interest/enjoyment (p=0.001) and the IMI-item perceived choice (p=0.010). We found no significant effects in the IMI-items man-machine relation and value/usefulness. Conditions promoting creativity (with or without visual display) were preferred compared to the ones not promoting creativity. An interaction effect of promotion of creativity and omission of visual display was present for training time (p=0.013) and training intensity (p<0.001). No differences in relative perceived training time, perceived effort, and perceived value among the four training conditions were found.

Conclusions: Promoting creativity in a visuo-audio-haptic or audio-haptic environment increases motivation in robot-assisted therapy. We demonstrated the feasibility of performing an audio-haptic music creation task and recommend to try the system on patients with neuromuscular disorders.

Trial registration: ClinicalTrials.gov, NCT02720341. Registered 25 March 2016, https://clinicaltrials.gov/ct2/show/NCT02720341.

Keywords: Audio-haptic display; Creativity; Intrinsic motivation; Music therapy; Robotic arm rehabilitation; Serious games; Stroke; User interface.

Fig. 1

ARMin arm rehabilitation robot. Additionally for this study, a keyboard was placed close to the participant’s left hand so that the space bar could be used as input device

Fig. 2

Screenshot of the tasks in V+ conditions with additional map of haptic elements: The visual environment displays sound icons (e.g., a bell or hands clapping) within sound zones (SZ), and a grey square that marks the center zone (CZ). SZs and CZ (wave-marked areas) are active zones where a vibrotactile haptic feedback is provided. The scene is limited by two haptic walls (HW, solid lines). HWs (dotted lines) limiting CZ are only turned on in the game phases where only the left or the right game zone is used. The red circle is the game “cursor”. All units are in centimeters

Fig. 3

Game rules for all four conditions. C+ conditions and C– conditions for all three game phases (i.e., active movement phase, listening phase, final phase) are presented. In C+ conditions, four successively activated sounds add up to a music piece (upper field, from top to bottom: claps, drums, guitars, marimba, …). In C– conditions (lower field), each consecutively activated sound is played alone

Fig. 4

Within subject corrected rating of subscale interest/enjoyment (left) and perceived competence (right) for visualization of the significant effect in promotion of creativity (C+ instead of C–) using linear mixed models. The black line indicates the mean of the performances the line is crossing. Each condition was played by a different number of subjects (n)

Fig. 5

Perceived size of value indicated by the desire to receive the music on mp3