Quantifying Learning in Young Infants: Tracking Leg Actions During a Discovery-learning Task

Motivating Spontaneous Infant Kicking Motions through Long Term Learning Utilizing a Robotic Mobile System

Our research investigates methods and systems to allow for early detection of motor impairment in infants and innovative interventions with the goal of improving longterm outcomes. A robotic baby mobile is utilized to motivate spontaneous kicking motions, which is used as a marker for predicting the potential of motor development delays. Our previous work investigated how the different stimuli modalities of a baby mobile can encourage infant kicking. We utilized a 3D camera system to detect the kicking motions, as well as recorded specific metrics of each kicking episode. In this work, we investigate the possibility of an infant having a preference of baby mobile stimuli that results in increased and sustained kicking motions. This preference is learned over multiple sessions with one infant and utilizes a Markov Decision Process to develop a policy.

Making the World Behave: A New Embodied Account on Mobile Paradigm

In this review article, we describe the mobile paradigm, a method used for more than 50 years to assess how infants learn and remember sensorimotor contingencies. The literature on the mobile paradigm demonstrates that infants below 6 months of age can remember the learning environment weeks after when reminded periodically and integrate temporally distributed information across modalities. The latter ability is only possible if events occur within a temporal window of a few days, and the width of this required window changes as a function of age. A major critique of these conclusions is that the majority of this literature has neglected the embodied experience, such that motor behavior was considered an equivalent developmental substitute for verbal behavior. Over recent years, simulation and empirical work have highlighted the sensorimotor aspect and opened up a discussion for possible learning mechanisms and variability in motor preferences of young infants. In line with this recent direction, we present a new embodied account on the mobile paradigm which argues that learning sensorimotor contingencies is a core feature of development forming the basis for active exploration of the world and body. In addition to better explaining recent findings, this new framework aims to replace the dis-embodied approach to the mobile paradigm with a new understanding that focuses on variance and representations grounded in sensorimotor experience. Finally, we discuss a potential role for the dorsal stream which might be responsible for guiding action according to visual information, while infants learn sensorimotor contingencies in the mobile paradigm.

In-Home Kicking-Activated Mobile Task to Motivate Selective Motor Control of Infants at High Risk of Cerebral Palsy: A Feasibility Study

OBJECTIVE

Children with spastic cerebral palsy (CP) have gait impairments resulting from decreased selective motor control, an inability to move the leg joints independently of one another, relying on excessive flexion or extension coupling across the 3 joints. Infants with white matter injury are at high risk of CP and have decreased selective motor control as early as 1 month corrected age. An in-home kicking-activated mobile task was developed to motivate more selective hip-knee control of infants at high risk of CP. The purposes of this study were to determine the feasibility of the in-home mobile task and to determine whether infants at high risk of CP and infants with typical development (TD) learn the association between their leg movements and mobile activation.

METHODS

Ten infants at high risk of CP based on neuroimaging and 11 infants with TD participated in this cohort study at 3.5 to 4.5 months corrected age. Each infant participated in the in-home kicking-activated mobile task for 8 to 10 min/d, 5 d/wk, for 6 weeks. Learning was assessed weekly based on an increase in the time that the infant demonstrated the reinforced leg actions when interacting with the kicking-activated mobile compared with spontaneous kicking.

RESULTS

With regard to feasibility, participation averaged 92% for infants at high risk of CP and 99% for infants with TD. With regard to learning, the group at high risk of CP demonstrated learning of the task for 2 of 6 weeks, whereas the group with TD demonstrated learning for all 6 weeks.

CONCLUSIONS

Infants at high risk of CP demonstrated learning of the kicking-activated mobile task but at a reduced amount compared with infants with TD. Further research is necessary to determine whether the kicking-activated mobile task has potential as an intervention to motivate more selective hip-knee control and improve walking outcomes of infants at high risk of CP.

IMPACT

This study investigated the feasibility of an in-home kicking-activated mobile task, a discovery learning task designed to motivate infants at high risk of CP to engage in the intensive task practice necessary to promote their learning abilities and selective motor control.

LAY SUMMARY

CP is a lifelong disorder of movement caused by abnormal development or early damage to the brain. If an in-home infant kicking-activated mobile task could be used to motivate certain types of age-appropriate leg movements of infants who are at high risk of CP, the task could help improve walking outcomes, which eventually could contribute to improving children's ability to participate in daily life. This study showed that infants at high risk of CP did learn the infant kicking-activated mobile task but at a much reduced amount compared with infants who are developing typically; so, this is a first step in determining whether the task has potential to motivate more age-appropriate leg movements in infants at high risk of cerebral palsy.

Sensorimotor Contingencies as a Key Drive of Development: From Babies to Robots

Much current work in robotics focuses on the development of robots capable of autonomous unsupervised learning. An essential prerequisite for such learning to be possible is that the agent should be sensitive to the link between its actions and the consequences of its actions, called sensorimotor contingencies. This sensitivity, and more particularly its role as a key drive of development, has been widely studied by developmental psychologists. However, the results of these studies may not necessarily be accessible or intelligible to roboticians. In this paper, we review the main experimental data demonstrating the role of sensitivity to sensorimotor contingencies in infants' acquisition of four fundamental motor and cognitive abilities: body knowledge, memory, generalization, and goal-directedness. We relate this data from developmental psychology to work in robotics, highlighting the links between these two domains of research. In the last part of the article we present a blueprint architecture demonstrating how exploitation of sensitivity to sensorimotor contingencies, combined with the notion of "goal," allows an agent to develop new sensorimotor skills. This architecture can be used to guide the design of specific computational models, and also to possibly envisage new empirical experiments.

Infant Discovery Learning and Lower Extremity Coordination: Influence of Prematurity

AIMS

Preterm infants at increased risk for neurodevelopmental disabilities, including cerebral palsy, demonstrate reduced selective leg joint coordination. Full-term infants demonstrate more selective hip-knee coordination when specific leg actions are reinforced using an overhead infant mobile. The purpose of this pilot study was to determine the ability of preterm infants to: (1) perform and learn through discovery, the contingency between leg action and mobile activation, and (2) demonstrate more selective hip-knee coordination when leg actions are reinforced with mobile activation.

METHODS

At both 3 and 4-months corrected age, ten infants born very preterm and with very low birth weight participated in 2 sessions of mobile reinforcement on consecutive days.

RESULTS

The preterm group at 4-months, but not 3-months, learned the contingency between leg action and mobile activation. Preterm infants at 4-months were separated into those that learned (n = 6) and did not learn (n = 4) the contingency. As a group, preterm infants at 4-months who learned the contingency, did not demonstrate more selective hip-knee coordination when interacting with the mobile on Day 2 as compared to spontaneous kicking on Day 1.

CONCLUSIONS

Preterm infants, as compared to full-term infants, may have difficulty producing more selective hip-knee coordination during task-specific leg action.

Lower limb pose estimation for monitoring the kicking patterns of infants

Monitoring the spontaneous kicking patterns of infants can give insight into their development. A computer vision based method for estimating the pose of an infant's leg from range images is presented in this paper. After some manual inputs for initialization, the range data associated with the infant is extracted. The method uses Robust Point Set Registration (RPSR) to fit an articulated model to the subject in every frame in the sequence, thus it provides the joint trajectories over time of the kicking kinematics. For validation, the method is used to track the articulation of a robotic humanoid that was programmed to kick in a fashion similar to an infant. Furthermore, the method is applied to a sequence collected from an actual infant and the resultant signal estimates are presented.