Mother-Infant Interaction Kinect Analysis (MIIKA)

Overview

An innovative methodology is developed to better understand parent-infant relationships. RGB-D sensors (such as Microsoft Kinect) give us the opportunity of online skeleton extraction based on the joints architecture of human bodies. These technologies provide automatic quantitative information of dyadic play, in order to get micro and macro features of the dynamic flow occurring during the interaction. This methodology looks at both behavioral features and objective measurement of spatial proximity and variations during free and structured interactions.

RATIONALE: Children developmental outcomes are known to be shaped by the early context of care and caregiver-infant interaction: an early sensitive caregiving environment provides an optimal emotional context for children's early brain maturation. Consequently, the multilayered and complex interactional processes occurring between infants and their mothers is of vital importance for healthy developmental trajectories. Although mother-infant communication is expressed and regulated using multiple modalities, only few studies focused on the role of body movements within the dyadic communication. Observational instruments - such as PICCOLO (Parenting Interactions with Children-Checklist of Observations Linked to Outcomes) - are useful to get the outstanding characteristics of the dynamicity and complexity of parent-infant relationships, but advances in technology allows us to analyze the flow of movements characterizing information exchanges in the bidirectional process of the parent-infant relationship. Spatial proximity is a core aspect of early dyadic relationships. The temporal and spatial reliability of automatic and computational methods, together with the decreasing costs of technology, are paving the way to the intersection of behavioral and technological techniques in both clinical and research fields. Nevertheless, to the best of our knowledge, kinematic full skeleton registration of mother-infant interaction in a free-play setting with children 18-36 months has not yet been explored both in at-typical and atypical development dyads. Indeed, previous studies assessed social interactions using automatic and computational methods, but mostly focusing on gesture and posture, gaze and mimics, quantity of body, facial and hand movements. SPECIFIC AIM 1: To identify spatial proximity features during free and structured dyadic interaction of typical development dyads and its association with mother and infant characteristics EXPERIMENTAL DESIGN AIM 1: mother-infant dyads will be asked to play within a specific perimeter that correspond to the detection area of Microsoft Kinect. The play session is divided into 3 phases: (1) free play. Some standard toys are positioned on the floor. (2) tower-task: mother and child are asked to play in order to build a tower with some cubes. (3) Empathy task: the mother is asked to pretend to hurt herself until a "stop" signal from the experimenter. Each session will be videotaped and data from Microsoft Kinect will be detected. Mothers will be asked to answer some questionnaires and expert coders will do behavioral coding with PICCOLO. Spatial proximity measures will be developed through algorithms able to extract, from Kinect row data, measures of approaches and separations of the dyad and mother and child's reciprocal contributions SPECIFIC AIM 2: to understand differences regarding the use of spatial proximity and reciprocal contributions of mother and infant both in typical and atypical dyads EXPERIMENTAL DESIGN AIM 2: children with neurodevelopmental disabilities and their mothers will be asked to play in the same setting explained in "experimental design 1". Data and proximity patterns that will emerge from typical development dyads will be compared to proximity patterns of atypical development children dyads, in order to understand how neurodevelopmental disabilities affect mother-infant interaction from a spatial and motor point of view. SIGNIFICANCE AND INNOVATION Although mother-infant communication is expressed and regulated using multiple modalities, only few studies focused on the role of body movements and spatial proximity within dyadic communication. This study wants to develop and test an innovative methodology based on RGB-D sensors applied to the setting of free parent-infant interaction. Automatic quantitative information of the dyad will provide micro and macro features of the dynamic flow occurring during the interaction. Thus, automatic quantitative information will supplement behavioral assessments, enriching these observations with objective measurement of spatial proximity and variations during free and structured interactions.

Conditions

Interventions

Eligibility

Locations (1)

Outcomes