Using Head-mounted, Eye-tracking to Compare Looking, Manual Exploration and Social Interaction During Active Locomotion

Using Head-mounted, Eye-tracking to Compare Looking, Manual Exploration and Social Interaction During Active Locomotion in Toddlers With Mild, Moderate and Severe Motor Delays: A Randomized Controlled Trial

The three purposes of this study are: 1) to determine the feasibility of using a head-mounted eye-tracker and examine the effects of the optimal mobility training on increasing visual and manual exploration in toddlers with different severity levels of motor delays, 2) to compare the effectiveness of providing optimal mobility training on outcomes across the ICF framework in toddlers with mild, moderate and severe motor delays, and 3) to determine the effects of optimal mobility training on the relationship among looking and manual actions, caregiver-child interactions and mastery motivation in toddlers with mild, moderate and severe motor delays.

Based on the power analysis from the preliminary results of the pilot study, we will recruit 72 toddlers with motor delays and further assign them to the mild, moderate or severe motor delays group. The participants in each group will be randomly assigned to either the ROC training program or conventional therapy (Control) program, resulting the following six training groups: the ROCStand(Mild) group (n=12), the ROC-Stand25(Mod) group (n=12), the ROCSit(Sev) group (n=12), Control(Mild) group (n=12), Control(Mod) group (n=12), and Control(Sev) group (n=12). The whole study duration will be 24 weeks, including a 12-week intervention and a 12-week follow-up. All programs will include 120 minutes/per session, 2 sessions/per week. Participants will continue their regular therapy during the whole study. The head-mounted cameras worn by the participants and caregivers will record the visual and manual behaviors for 40 minutes/per week during intervention. Standardized assessments are provided for a total three times, including the time before and after the intervention and in the end of the follow-up phase. Assessments include body function, mobility, psychosocial function, family perception and participation. The findings of this study will provide guidelines for arranging the training environment depending on each child's and family's needs.

Although environmental and task modifications are powerful methods used to affect function in rehabilitation and frequently used by the therapists, these modifications are typically introduced using a trial and error approach. A better understanding of the fits between the child and the environment may help therapists develop a more structured approach to manipulating environmental and task characteristics in order to enhance the emergent actions of children. The important factors required to achieve positive effects of ride-on car (ROC) training include the"fit"between a child's characteristics, the modified ROC, and the physical and social contexts.

Based on the findings of previous randomized controlled trials (RCTs) (NCT03707405), children with mild motor delay may gain the most from using the ROC-Stand style for 70 minutes (ROC-Stand) while those with moderate motor delay may gain the most with a 25-minute standing posture and a 45-minute sitting posture (ROC-Stand25). Children with severe motor delay may benefit more from the ROC training in a sitting posture (ROC-Sit), which may provide moderate challenges as an optimal mobility training program. Thus, the heterogeneity of the participants regarding severity level of motor delays and environmental considerations may be the factors contributing to the different treatment effects.

A new technical challenge occurs to record and capture young children's visual and haptic exploration. Infants cannot attend to the entire visual field captured by the action camera and their attention constantly shifts between areas of the visual field. It is difficult to differentiate what the targets the child actually fixates at or interacts with. To allow the direct observation of infants' spontaneous eye movements during self-initiated locomotion, Dr. Adolph have used a wireless, head-mounted eye-tracker (Positive Science, LLC) which could measure infants' visual behaviors during free, unfettered locomotion. This novel methodology includes a lightweight and comfortable headgear, transmitter, and battery pack which can record infants' eye gaze. The analysis can focus on where infants look during self-initiated locomotion, object exploration and in response to caregivers' vocalizations. The combination of the ROC training with this head-mounted eye-tracker will allow us to further track what visual and haptic information is obtained during the early PMD use. It is helpful for understanding the influences of environmental factors on psychosocial function in toddlers with motor delays.

To provide a full investigation of the effectiveness of optimal mobility training, this study proposes to compare the effects on toddlers with mild, moderate and severe motor delays. Additionally, a wireless, head-mounted eye-tracker will be applied as the novel method to analyze children's looking, manual behaviors and social interaction during the training phase.

The specific aims of this study are: 1) to determine the feasibility of using a head-mounted eye-tracker and examine the effects of the optimal mobility training on increasing visual and manual exploration in toddlers with different severity level of motor delays, 2) to compare the effectiveness of providing optimal mobility training on outcomes across the ICF framework in toddlers with mild, moderate and severe motor delays, and 3) to determine the effects of optimal mobility training on the relationship among looking and manual actions, caregiver-child interactions and mastery motivation in toddlers with mild, moderate and severe motor delays.

Method Study Design: A multiple pretest-posttest control group design will be applied. On the basis of our previous pilot findings in toddlers with mild, moderate and severe motor delays, 6 groups will be involved in this project: using a ride-on car in a standing posture for driving for children with mild motor delay (ROC-Stand(Mild)), conventional therapy for those with mild delay (Control(Mild)), a 25-minute standing posture and 45-minute sitting posture for driving for those with moderate motor delay (ROC-Stand25(Mod)), conventional therapy for those with moderate delay (Control(Mod)), a sitting posture for those with severe motor delay (ROC-Sit(Sev)), and conventional therapy for those with severe delay (Control(Sev)). Based on the previous RCTs, the study duration for each participant is 24 weeks, including a 12-week training phase (intervention) and a 12-week follow-up phase.

Participants: 72 participants aged 12 months to 36 months old with motor delays will be recruited in the study and determined the severity level of motor delays based on the cutoff values of Bayley-III MC scores (mild: 70-84; moderate: 55-69; severe: <55). 24 toddlers with mild motor delays will be in one group, 24 toddlers with moderate motor delays in another group and the other group has 24 toddlers with severe motor delays. Participants in each group will be randomly assigned to a ROC training group or conventional therapy (Control) group. There will be 6 groups, including the ROC-Stand(Mild) group (12 toddlers), the Control(Mild) group (12 toddlers), the ROC-Stand25(Mod) group (12 toddlers), the Control(Mod) group (12 toddlers), ROC-Sit(Sev) group (12 toddlers) and Control(Sev) group (12 toddlers).

Recruitment: The participants will be recruited from self-referrals, health care practitioners, or the hospitals in Taipei or Taoyuan where toddlers with motor delays are receiving outpatient rehabilitations. The research team will initially post flyers describing the study at clinical settings and contact the therapists in the clinical settings to introduce the goals, criteria and general procedure of the study. Parents/guardians will obtain information about the study through the flyers and their therapists. The research team will contact the parent/guardian to explain study details and provide them the opportunity to ask questions. Children of parents/guardians who provide informed consent will participate in the study.

Procedure: Before randomization, the participants will receive pre-intervention measurements, including standardized assessments of developmental abilities, the evaluations of environmental factors and parental-perception questionnaires. An independent, licensed OT or PT who is blinded to the study purpose and not involved in the training programs will complete the pretest and posttest assessments and questionnaires in a testing room at Chang Gung University. These assessments will occur on three occasions: before and after the 12-week training phase, i.e., intervention (T1 & T2) and the end of the 12-week follow-up phase (T3). After the pretest, participants are arranged to the group of mild motor delays, moderate motor delays or severe motor delays based on the cutoff values of Bayley-III MC scores. After allocation, participants will be randomly assigned into the ROC training or conventional therapy groups, resulting in 6 groups. All the participants will start the training programs in a training room and certain public space in the university. During the 12-week intervention, the locomotor experience of eye gaze and the accompanying manual actions, and social interactions for the 6 training groups will be recorded by a wireless, head-mounted eye-tracker worn by the participants for one 2-hour session/per week at the university. An activity log used in the previous studies will also be applied to record the training conditions and the caregiver's feedback on the training programs every week during the 12-week intervention. The activity log will be mostly qualitative data that involves caregivers' descriptions. For the 12-week follow-up phase, all participants will not receive any training program from the research team. They only have their regular therapy from their own hospitals or clinics. The information of regular therapy, including the type, the frequency and the dosage, will be collected at T1, T2 and T3.

Intervention: The therapists who only involve providing the training programs will ask caregivers to identify goals and measure progress using goal attainment scaling (GAS) at T1, T2 and T3 time points for the 6 groups. Two independent licensed therapists (OT or PT) who will not involve the administration of assessments will provide the ROC training or conventional therapy programs, respectively. One training provider is responsible for the ROC-related programs which require the therapist working with the caregivers based on the coaching technique described in the FCS model. The ROC involves a standing, a combined or a sitting posture depending on their severity level of motor delays. The training principles are similar to those applied in the previous studies of ROC training in various environments. The other therapist is responsible for the conventional therapy which is a child-centered, therapist-directed program. Participants will have a necessary break depending on their endurance during training. In addition, all participants in the 6 groups will wear the eye-tracking headgear to record the looking, manual and interactive behaviors for one session/per week during intervention. All participants in the 6 groups will continue their regular therapy from their own hospitals or clinics throughout the 24-week duration of the study, including physical therapy, occupational therapy, and speech therapy.

Follow-up: This period will involve a 12-week phase following the above treatment programs; during this time no treatment programs will be delivered to the participants except for their own regular therapy.

Instrument and Procedure: To obtain the data of locomotor experience, an objective measure of a head-mounted, eye-tracker and a qualitative measure of activity logs will be applied during the 12-week intervention phase. Standardized assessments as the quantitative measures across the ICF framework will be administered at T1, T2, and T3 in a testing room of the university.

Descriptive statistics (frequency, means, standard deviations) of each behavior will be calculated. Kolmogorov-Smirnov will be used to examine whether the data follows a normal distribution. To compare the baseline characteristics of the major 3 groups categorized as mild, moderate and severe motor delays, one way ANOVA (for data with normal distribution) and Kruskal-Wallis analysis of variance (for data with non-normal distribution) will be conducted. Data will be analyzed based on an intention-to-treat analysis. For those data collected during the 12-week intervention, descriptive statistics (frequency, means, standard deviations) of looking, manual actions and interactive behaviors will be calculated and repeated measures analysis of variance (group [6] × time [12]) will be used to compare the effects on each exploratory behaviors among the 6 training groups during intervention. A repeated measures analysis of variance (group [6] × time [3]) will be employed to evaluate the treatment effects on the secondary outcomes within and among the 6 groups at T1, T2, and T3, followed by a post-hoc analysis using Bonferroni test to determine between which groups the differences occur. To clarify the relationship between other relevant factors and psychosocial outcomes, we will consider two co-variates in the analysis based on previous research: MC scores from the Bayley-III and caregiver-child dysfunctional interaction (obtained from PSI). Last, we will apply a coefficient test for analyzing the relationship among the social function, looking/manual actions and social interactions, and mastery motivation during the intervention phase in each group. The scores of social function and mastery motivation will be based on the mean change scores from T1 to T2. The mean frequency of total looking/manual actions and social interactions will be calculated from the coding results of the 12-week intervention.

Inclusion Criteria:

1. motor delays that resulted in motor impairments preventing independent walking (standard deviation [SD] <-1.5, assessed by the Chinese Child Development Inventory via a pediatric physician)
2. being able to stand independently for two seconds or tolerate standing with support for 10 minutes
3. being able to reach objects with either one or both hands
4. 69 to 103 cm height and 7 to 18 kg weight
5. parents agreeing to provide consent for their child's participation in the training program

Exclusion Criteria:

1. children with severe sensory impairments (including blindness or deafness)
2. exceeding the height or weight criteria
3. parents/guardians who were unable to make the time commitment for the program and children with severe emotional disorders resulting in strong reactions and causing possible harms.