TRIAD receives grant funding to study precursors of challenging behaviors, evaluate use of telemedicine for autism assessment

The Vanderbilt Kennedy Center’s Treatment & Research Institute for Autism Spectrum Disorders (VKC TRIAD) was recently awarded funding on two major multiyear, multi-site research grants, both with the aim to use technology to improve the lives of individuals with autism spectrum disorders (ASD) and their families.

Wearable technology and identifying pre-cursors of challenging behaviors

The National Science Foundation (NSF) awarded a Smart and Connected Health Grant* to VKC researchers Nilanjan Sarkar, Ph.D., (Principal Investigator), Amy Weitlauf, Ph.D., (Co-PI, TRIAD associate director of Research) and John Staubitz, M.Ed., BCBA (TRIAD associate director of Behavior Analysis) to develop and study the use of wearable technology to identify precursors of challenging behaviors in individuals with disabilities, in an effort to mitigate those behaviors and improve safety.

“Many children and adolescents with intellectual and developmental disabilities show severe problem behaviors that present safety risks and interfere with living a fully included life in the community,” said Amy Weitlauf. “Effective gold-standard behavioral approaches exist that help us better understand those behaviors, such as when and why they are most likely to happen. But those approaches require a lot of human effort, time, and training, and sometimes involve risk. The goal of this work is to figure out if wearable technology and machine learning can help make behavioral approaches safer, faster, and easier to implement in the community.”

Vanderbilt will partner with researchers from North Carolina State University over the course of four years on this project, which first began a few years ago as a VKC U54 Signature Project pilot study.

“We are planning on creating new machine learning models. These models will take several physiological signals such as heart rate, skin sweating, body movement, and eye gaze as input, and then fuse those signals into an actionable output that indicates a precursor of problem behaviors may be occurring,” said Nilanjan Sarkar.

“Ultimately, we hope to develop predictive models based on these signals that will alert parents and caregivers – and, in the future, the individuals undergoing treatment themselves – to an impending problem behavior, so that therapist-informed strategies can be used to lessen its impact or occurrence.”

Sarkar and his team will develop wearable tech that can alert therapists to physical changes, like increased heart rate, that the therapist cannot perceive. Other data will be collected through a custom app along with therapist observation notes, all of which is then fed into a complex machine learning algorithm to create a personalized model for the patient.

“Our hope is that these technologies will give therapists another tool that will improve the practice that they’re already using. It won’t take the place of a well-trained behavioral therapist, but it could enhance what they are able to measure and inform how they approach their patients to maximize the safety and efficacy of how time is spent,” said Weitlauf.

Vanderbilt and NC State will work together to refine the technology using input gathered from patients and therapists to create a product that is low-cost and efficient while also being comfortable and subtle. Once the tech is streamlined, Staubitz will use his background as a Board-Certified Behavior Analyst to lead the community implementation and evaluation portion of the research.

“A model for reliably predicting behavioral escalation would allow service providers to conduct therapeutic sessions that are safer, more dignifying, and that enhance the efficiency and independence of the child or adolescent,” said Staubitz. “This current study will focus on the impact of the technology on the safety and efficiency of more traditional intervention sessions, but the technology design is intended to make it feasible for this technology to be extended into the classroom, home, or community if beneficial.”

“This work is especially exciting because it represents years of collaboration between TRIAD researchers, behavioral and community systems experts, and our engineering partners,” added Amy Weitlauf. “It’s a fantastic opportunity that would not have been possible without these multidisciplinary collaborations.”

Bringing autism assessment into the home

VKC investigator and TRIAD executive director Zachary Warren, Ph.D., recently received a five-year multi-site R01 research grant** to study the use of telemedicine for assessing autism in toddlers, using a tool that lets clinicians observe parent-child interactions within the child’s home. Warren serves as Principal Investigator, with assistance from Amy Weitlauf, Ph.D., Laura Corona, Ph.D., and Liliana Wagner, Ph.D., BCBA.

Prior to the COVID-19 pandemic, Warren and his team already had NIH funding to study and develop a telehealth assessment tool (then named TELE-ASD-PEDS) with the goal of increasing access to services for families living in more rural, medically underserved parts of the state. In that study, families came to research labs at Vanderbilt for assessment. Psychologists interacted with families through a computer, guiding parents to administer certain activities and tasks with their child. The clinician would observe the child’s behaviors remotely and then score the results to make an assessment.

When Tennessee went into pandemic lockdown in March 2020, clinical visits came to a grinding halt. In a field where early assessment leads to better developmental outcomes for a child with disabilities, the idea of an indeterminable delay in diagnosis was problematic. Warren felt confident enough in the tool to lead multiple online workshops on the TELE-ASD-PEDS (now abbreviated to TAP) with thousands of clinicians across the country, and feedback was overwhelmingly positive.

This new R01 grant will allow the TRIAD team to study the use of the TAP, using tele-assessment to bring the psychologist to the child in their own home. Research will be conducted at the VKC as well as at the University of California-Davis, with Sally Ozonoff, Ph.D., serving as PI.

“We have been struggling for decades to design better systems of care for traditionally underserved and geographically isolated communities,” said Zachary Warren. “The recent TELE-ASD-PEDS/TAP work highlights ways in which we can do more to meet families’ needs in their own backyard.”

“This grant came about because there are things that we still don’t know about tele-assessment. Even though we have this tool that we feel positively about, we still need information to validate the use of the TAP in a direct-to-home setting,” said Laura Corona. “There are three pieces to this: one, can we use tele-assessment to make an autism diagnosis that is reliable and consistent with the diagnosis from a traditional face-to-face evaluation? Second, how would families respond to that type of assessment? Is tele-assessment meaningful to families; do they find it to be an acceptable method of assessment? And the third, how does tele-assessment set families up for accessing early intervention and the resources that they need for their children? That’s a really big question, because one of our main goals in using telemedicine is being able to reach families who face barriers to traditional assessments.”

The TAP is a series of activities that a psychologist guides a parent to complete with their child, in their own home. Activities include, for example, playing together with toys, calling the child’s name, and offering a snack. Families are connecting with the observing psychologist using their own technology such as a home computer with webcam or, more frequently, a cell phone.

“Our preliminary work is telling us a couple of things. One is that the home environment is more comfortable for some families. Providers like that they are able to see to see the child in their home environment, which is more natural than bringing them into a clinic environment,” said Corona. “The home environment is also different in that we have less control over it. In the clinic, we provide the technology, the toys, the space. In the home environment, there are often more family members present, the family uses the items that they have, and they use their own technology. Those things can all work well. They can also present challenges, if we don’t have a good Wi-Fi connection, or if someone’s cell phone dies. If they don’t have a lot of toys, or if they have tons of toys that are maybe a bit distracting. One of the purposes of this work is to really look at all those factors.”

The overall aim of this research is to better serve families from racial or ethnic minority backgrounds, families living in poverty, and families from medically underserved, under-resourced communities. As such, provisions can be made for families who don’t have easy access to technology or internet connectivity.

“It is estimated that 70 to 80 percent of American adults have access to a smartphone with connectivity and internet access sufficient for a tele-assessment, including those in low income, rural, and minority communities,” said Liliana Wagner. “For those families who do not have access to the needed technology, we will arrange through our early intervention and primary care partners for the telehealth visit to take place in a community setting close to home, such as the health department, an early intervention center, or a health clinic. We have done this successfully in preliminary studies, in those rare instances when that barrier is present.”

“I think that the past year and a half has taught us that telemedicine is possible and useful and also beneficial, not only for families who live far away, but also for families who want the convenience and comfort of being in their home,” added Laura Corona. “There are so many reasons that families might find telemedicine to be a really good option, so it is exciting to have a tool that will let us do this kind of evaluation.”

For more information on the VKC Treatment & Research Institute on Autism Spectrum Disorders, visit TRIAD.vumc.org or call (615) 322-7565.

*National Science Foundation, Award Number 2124002: SCH: Enhanced detection of impending problem behavior in people with intellectual and developmental disabilities through multimodal sensing and machine learning. Funding Opportunity: NSF 21-530 Smart Health and Biomedical Research in the Era of Artificial Intelligence and Advanced Data Science.

**Department of Health & Human Services, Award Number 1R01MH127228-01: Addressing disparities in ASD diagnosis using a direct-to-home telemedicine tool: Evaluation of diagnostic accuracy, psychometric properties, and family engagement. Mental Health Research Grant.