Sensorimotor processes underpinning imitation learning of biological motion

• Main Author: Dutoy, C. A.
• Other Authors: Hayes, S. J. ; Bennett, S. J. ; Elliott, D.
• Published: Liverpool John Moores University 2018
• Subjects: RC1200 Sports Medicine
• Online Access: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.743398

The aim of the present thesis was to examine the way in which biological motion is coded and imitated during imitation learning by improving upon methodologies currently used in the literature to examine imitation of underlying movement kinematics. Across four experiments, imitation of the kinematic structures of biological and non-biological motion models was examined to investigate the processes involved in imitation learning. The purpose of the first experimental chapter, Chapter Two, was to examine the way in which biological motion kinematics were coded during imitation learning by establishing whether imitation of biological motion kinematics was a function of lower-level visuomotor processing or top-down attentional modulation. Results showed that not only were imitations of typical and atypical biological motion different, but both models were imitated as accurately during spatially incompatible trials as compatible. Accurate imitation of spatially incompatible atypical biological motion confirmed biological motion coding is a function of lower-level visuomotor processing. Following results from Chapter Two, Chapters Three, Four and Five assumed lower-level visuomotor processing of biological motion and were designed to further examine whether this lower-level visuomotor processing of biological motion was modulated by top-down attentional factors (e.g. end-state-targets, visual attention, social primes). The first of these top-down modulations was included in Chapter Three, which examined the influence of end-state-targets on biological motion coding during imitation learning. Although kinematics was not modulated by end-state-targets, movement time was less accurate when end-state-targets were present, which suggests that lower-level and top-down processes operate together during the processing of visual information during imitation learning. In addition to end-state target modulation, imitation data further confirmed the coding of atypical biological motion by demonstrating differences in imitation of two relatively similar atypical biological motion models (atypical17 and atypical26). The top-down attentional factor examined In Chapter Four was visual attention, which was measured by recording eye movements during observation of the model stimuli. Analysis of eye movements demonstrated that visual attention was directed towards the model throughout the entirety of the observation phase during trials where end-state-targets were both present and absent. As goal-directed eye movements were not made during observation of the models, results suggest that the kinematic data contained within each of the models was observed and consequently featured in the representation formed for motor execution. Chapters Two, Three and Four provide a fundamental understanding of how biological motion is coded during imitation learning by using robust protocol that improves upon the validity of those used in the current literature and specific modulations that discredit significant top-down modulatory explanations for biological motion coding. The way in which biological motion coding occurs in neurotypicals (no neurologically atypical patterns of thought or behaviour) is important when trying to understand where deficiencies in those with intellectual disabilities occur. The intellectual disability most closely associated with the current thesis is autism, where deficiencies in imitation are suggested to be linked to social components. Therefore, to establish a foundational understanding of how social context influences neurotypical imitation, Chapter Five examined the influence of social primes on the coding of biological motion. Results showed that social primes modulated the accuracy of imitation, where peak velocity was more like those of the models following observation of an anti-social prime. In addition, observation of both the pro- and anti-social primes was shown to reduce the variability of imitation relative to observing no social prime at all. These findings demonstrate that social primes are being coded and incorporated into the motor output such that both the accuracy and consistency of imitation of biological motion are modulated. Together, the results presented in the current thesis demonstrate imitation of novel, atypical biological motion is a function of complimentary lower-level and top-down processes that facilitate the coding of both underlying kinematics and environmental context.