Shedding new light on autism spectrum disorder

Autism spectrum disorder (ASD) is among the most widely known—thanks in part to depictions in popular culture of individuals with ASD—and most perplexing of disabilities. According to the U.S. Centers for Disease Control, one in 54 children is now diagnosed as being on the spectrum, with boys four times more likely than girls to be diagnosed. The number of people known to have ASD has risen dramatically in the past 40 years.

For Michael Alessandri, executive director of the University of Miami-Nova Southeastern University Center for Autism and Related Disabilities (UM-NSU CARD) and professor of clinical psychology at the University of Miami, the more we discover about the causes and effects of ASD, the more of a moving target it becomes. He said that understanding the areas of differentiation on the spectrum is the first step to helping individuals on their journey with ASD, a complex, lifelong condition. “We’re very invested in research,” he said. “And I'm most interested in research that is designed to make people's lives better, that yields some kind of benefit to the people we are taking care of.”

Understanding the biological pathways that lead to ASD

UM-NSU CARD is a clinical site partner in the Simons Foundation Powering Autism Research for Knowledge (SPARK), the world’s largest autism research project, which seeks to identify genetic markers for the disorder. SPARK aims to collect DNA samples from 50,000 family “trios”—children with autism and both parents—from across the country, including 700 from UM-NSU CARD’s client network. SPARK researchers study the samples to identify the genetic markers associated with autism, which in turn will advance understanding of how the different markers function in different biological pathways and what interventions will be most effective.

“Fundamentally, autism is a neurobiological disorder caused by genetic abnormalities that affect the way the brain processes information,” Alessandri said. “For a long time, we were looking for an ‘autism gene’ as a cause and now that search has turned out to be a search for causes. And it is a spectrum—you have kids who have great language, great cognitive ability, and are somewhat socially motivated; then you have kids and adults who are much more intellectually impaired, nonverbal or minimally verbal, and really compromised in their ability to live their daily lives; and you have everything in between. And for each individual, there may be thousands of neurobiological, genetic pathways that are all different across different people.”

“I think we are in a place where we clearly know more than we did 25 years ago, even 10 years ago, but I think we’re probably still 10 to 15 years away from a really deep understanding of the nuances within the spectrum,” Alessandri said. “We have over 150 candidate genes that seem to be linked to autism, so I think we are closing in on the biological markers that cause autism. But with 150-plus candidate genes [and] multiple biological pathways, you get all different kinds of outcomes.”  

Technology-driven diagnostic and therapeutic approaches

As Alessandri and his colleagues in the Department of Psychology attest, early diagnosis and intervention are critical. “If you can diagnose a kid early—as early as 12 to 24 months—and get them into high quality intervention, you absolutely change the trajectory of their lives,” Alessandri said.  

ASD is defined behaviorally, rather than biologically, and that definition has evolved significantly in the diagnostic manuals that Alessandri and his clinical colleagues work with. “There are no blood tests, there’s no EEG or MRI, there’s nothing biological that we can use to diagnose autism,” Alessandri said. “You really have to diagnose it based on the observable behaviors you see in the client. And it is a moving target.”

Alessandri and psychology professor Daniel Messinger have teamed up with other colleagues in psychology, as well as those in electrical and computer engineering and physics, to develop more objective measures of autism symptoms. Messinger is a lead investigator in Behavioral Imaging of Autism for Science (BIAS). This project, supported by the University’s Clinical and Translational Research Institute, uses special glasses to measure social-communication behaviors, such as eye contact, smiling and other facial expressions, and language, that are key indicators of autism in young children.

“One characteristic that children with autism have is they make less frequent eye contact. And we’ve found that the more a child looks at the examiner, the less severe their symptoms. The idea of BIAS is to develop some objective measures of [these behaviors],” Messinger said. “We needed a technology that tells us how much a baby is looking at an examiner as they assess whether or not the baby has autism.” Using glasses with a small camera built into the bridge, the examiner can record the baby’s face as well as capture audio of the assessment. Messinger worked with colleagues in engineering who used machine learning to analyze the data and shed light on the diagnostic process.

Technology has also helped with advances in therapeutic interventions for people with ASD. The standard of care for autism therapy is Applied Behavior Analysis (ABA), which seeks to teach skills and reduce behavior issues that interfere with learning and social functioning. ABA has its shortcomings: it is expensive and labor-intensive, and the interventions generally take place in highly controlled environments that don’t equate to mastering life and job skills in the real world, which adults and adolescents with autism need.

A multidisciplinary team of University of Miami researchers is exploring how augmented reality technology and artificial intelligence (AI) can be used to resolve these shortcomings. Anibal Gutierrez, research associate professor of psychology, director of the Intensive Behavioral Intervention Services (IBIS) Clinic and the master’s program in Applied Behavioral Analysis in the Department of Psychology, and associate director of UM-NSU CARD, and Mohamed Abdel-Mottaleb, professor and chair of electrical and computer engineering, are part of a team that received a University Laboratory for Integrative Knowledge (U-LINK) grant for their project Augmented Reality for Applied Behavior Analysis—Evaluations, Creations, and Applications.

As Gutierrez explains, the project aims to put individuals “in virtual environments that are hard to replicate the in real world because of safety concerns, like placing trainees on simulated streets, or feasibility. One of the projects we are starting with is making a hamburger. This is where the technology steps in—a fully immersive virtual reality environment that walks someone through the process of making a hamburger.  Then we would use an augmented reality headset with an overlay that gives visual and auditory cues as the person works.”

Abdel-Mottaleb explained how AI will play a significant role in the augmented reality training. “The purpose of AI is twofold. The first is to recognize objects on the scene, and the second is to tune the cues given to the person toward that specific person. What we are trying to achieve is to build a virtual reality system integrated with AI, both for object recognition and for creating cues that are adapted to individual users,” he said.

“With augmented reality, you can glow certain objects, point with arrows, and reorder how you present the skills,” Abdel-Mottaleb said. “By examining how these work for a given child, you can [through AI] tune the system toward that child and adapt for each new child. Ultimately the best outcome would be a system, including hardware and software, that can be used by anyone, anytime. It would be cheaper and it would make the treatment reachable for anybody.”

The team hopes to begin testing the system on student volunteers later this year. In the meantime, Abdel-Mottaleb stresses the importance of a multidisciplinary approach to a challenge as big as ASD. “There is no one person who knows everything. So, for [the U-LINK] project, we need the domain knowledge coming from psychology. We need someone who can understand the interaction between a person and machine. And then we need the engineering part in order to be able to build and integrate the AI system.”

“It’s a journey”

For Alessandri, the autism research—“paramount, wonderful, complicated, frustrating” research—is ultimately about alleviating suffering in people with ASDs and helping them realize their full potential. For him, and others in the field, it’s not an easy road, but it is a rewarding one. “It’s a journey, and I’ve been on it a long time,” he said. “But we have a long way to go. I’m primarily a service provider, clinician, academic, but all of us doing research in autism are exceptionally motivated to find answers.”

Related Stories