A Smart Toy for Motor Function and Early Assessment of Children With Possible Autism (Tangiball)

"Tangiball": Acceptability of a Smart Toy Tool for the Clinical Detection of Sensori-Motor Dysfunction in Children With Possible Autism Aged 2-5

National Institute for Health Research, United Kingdom, Brighton & Sussex Medical School, Nottingham Trent University, University of Cambridge, Council for Disabled Children, Autistica

Autism diagnostic referrals across the UK have doubled within the last five years. Covid-19 has further delayed diagnostic pathways, while innovation has not kept pace in assisting clinical teams with low cost, fast, and unobtrusive pathways to shorten waiting times for families of children with possible Autism. Sensor-based technology offers a potentially cheap, small-scale, and unobtrusive way of collecting data while children interact seamlessly with smart play objects and toys that allows a clear comparison with neuro-typical groups or children.The present exploratory group intervention (intervention, interview, focus group, and questionnaire) will map out and investigate clinical interaction through the use of "Tangiball"-a new low-cost smart toy that is highly reliable and sensitive-that has the potential to significantly reduce the timing of the diagnostic process in young children with possible Autism. The "Tangiball" records user speed and accuracy of movement, which aligns with variability between neuro-typical children and children with Autism, but is it an acceptable clinical tool for diagnostics and play?

Sensor-based technology offers a potentially cheap, small-scale, unobtrusive, and data rich capture environment. Harnessing data capture whilst children interact seamlessly with play objects and toys, will enable large-scale data insight into the active behaviour of children, allowing clear comparisons with neurotypical groups. Data comparison can be modelled on diagnostic manual descriptions, and would enhance pathology descriptions. Smart systems can provide diagnostic evidence that can be compiled for children with disabilities and be appropriately compared to neuro-typical baseline data. Borderline diagnosis and confusion can be avoided, as collected data can then be used in addition to observable report. This data elicits new information about the way children with Autism and associated developmental disorders interact with their environment on a micro-level. Children with Autism when using smart objects show slow movement, where deceleration during grasping reaching and placing activities takes longer than neurotypical peers, the amplitude of reach speed is at its peak higher than typical peers, maximum acceleration through space is slower, and the time taken to reach speed peak is higher such that speed and accuracy can account for 76% of all variance between children with Autism and typical peers. Children with Autism therefore experience qualitative differences in the way they navigate their own body and interact with objects within their own environment. Stereotyped or repetitive motor movements, hyper- or hypo-reactivity to sensory input, and unusual interests in sensory aspects of objects, remain central to Autism diagnosis. Atypical sensory and physical behaviour is estimated in 92% of children with ASD and 67% of children with SEND equating to one million affected school children. Diagnosis and clinical interaction of children with possible autism currently relies on the use of subjective observational inventories (SOIs) such as ADOS, whilst engaging children in unfamiliar surroundings, against the backdrop of dwindling resources. Autism assessment referrals across KSS have doubled within the last five years. Covid-19 has further delayed pathways. Innovation has not kept pace in assisting clinical teams with low cost, fast and unobtrusive pathways to shorten waiting times. As clinical interactions with younger children nearly always utilise toys to improve engagement, this Smart Toy solution enables automated data capture-otherwise lost-by embedding sensors (e.g. RIFD, IMU Bluetooth) in toys. Accurate data capture during object interaction has potential to significantly reduce assessment time-a recognised problem highlighted in the NHS Long term plan. "Tangiball" is a new low-cost smart toy that is highly reliable and sensitive that has the potential to significantly reduce the timing of the diagnostic process in young children with possible Autism. The "Tangiball" records user speed and accuracy of movement, which aligns with variability between neuro-typical children and children with Autism. The present exploratory group intervention study (intervention, interview, focus group, and questionnaire) will map out and investigate clinical interaction through the use of sensors that can tell the investigators how children are interacting with objects, and how this diverges from neurotypical children. The project aims to establish the "Tangiball" as an acceptable tool for clinical staff, children, and parents during clinical interaction; that clinical interaction with a Smart Toy enables useful data to be gathered that can support the diagnostic process in young children. With few smart devices embedded into everyday clinical practice this project completes and trials an Internet of Medical Things Smart Toy for use in child development clinics, to capture child interaction with objects and assist in profiling. This can help determine appropriate pathways to care. Thirty children (n=15 with a formal diagnosis of Autism Spectrum Disorder, n=15 neurotypical children with no prior referral to a child development clinic), aged 2-5 years will be recruited over a period of 11 months via clinical teams and through schools. The intervention will run either in clinic or in children's homes. Children will play with the Smart Toy system during a 20 minute session. Children will be matched according to age, and score on the four-strand "Hands-On" assessment tool. Parents and children will complete a User Experience Questionnaire [UE-Q] which assesses user acceptability and experience of the Smart Toy. Sessions will be recorded using an SD card within the Smart Toy which records information on user speed and accuracy of placement of the detachable segments of the toy when placed into the main hub of the toy. Ten clinicians (comprising SALTs, Psychologists and Paediatricians) will be interviewed to explore themes around smart toy use and acceptability. Data produced will explore the acceptability and optimisation of triage using a novel, low-cost device that is highly reliable and sensitive, including physical and behaviour profiling to increase timely accurate identification of strengths and weaknesses in children in the early school-age, and pre-school population (2 to 5 years) where there is a current pressure on identification.

To enable baseline scores comparison, children will be matched according to age and score on the four-strand "Hands-On" assessment tool which takes 30 minutes to complete. Children will then play with the Smart Toy during a 20 minute session. Sessions will be recorded using an SD card within the Smart Toy which captures information on user speed and accuracy of placement of the detachable segments of the toy when placed into the main hub of the toy, allowing comparison between neuro-typical children and children with Autism. The intervention will run either in clinic (preferred option) or in children's homes-and in this case will be sent to them via post with the Standard Operating Procedure which can then be employed when children are taking part in the intervention via online video link (not recorded).

Data recording of movement on X, Y, Z axis (placement of jack, duration of placement within tangiball)

11 months or until recruitment target for intervention phase is reached (including 10 months recruitment phase)

Parents and children will complete a User Experience Questionnaire which assesses user acceptability and experience of the Smart Toy. This questionnaire asks a response to two binary choices on a 7-point likert scale e.g. "Annoying - Understandable" and so will give numerical response where the higher the score, the better the response. Deploy and assess acceptability of the Smart Toy.

11 months or until recruitment target for intervention phase is reached (including 10 months recruitment phase)

Baseline scores of children collected by means of "Hands-On" assessment will be compared to Accelerometer outputs.The "Hand-On" tool establishes scores across four dimensions in the 3-8 age group of obtaining objects, practical use of objects, repensetattional use of objects, and social use of objects. Each dimension is scored according to sub-categories on a scale of "Mastered" (score = 2), "Emerging" (score = 1) or "Not present" (score = 0). Each dimension is then calculated as a subtotal before a grand score out of 78 (also calculated as a percentage). The higher the score the better the child's interaction with objects across the four dimensions.

11 months or until recruitment target for intervention phase is reached (including 10 months recruitment phase)

Inclusion Criteria: Intervention phase: child aged 2-5 years and either: diagnosed with Autism is a neurotypical child with no prior contact with a Child Development Clinical Service for social communication problems Focus groups, interviews, surveys: either a parent of a child diagnosed with autism, or clinician who is involved in autism diagnosis Exclusion Criteria: Intervention phase: neurotypical child has prior referral to Child Development Clinical Service Child child is outside of age range child not under the care of Child Development Clinical Service Focus groups, interviews, surveys: parent not of child with Autism, or a clinician not involved in autism diagnostic services