As we prepare to Light It Up Blue on April 2, please join us in celebrating the contributions of our long-time science advisor and pioneering autism tech researcher Matthew Goodwin.
Follow the news at #LIUB
Post by Matthew Goodwin, an advisor to Autism Speaks’ Innovative Technology for Autism initiative and member of our Scientific Advisory Board. Dr. Goodwin pursues autism research at his Computational Behavioral Science Lab at Northeastern University’s Bouvé College of Health Sciences & College of Computer and Information Science. He is also a visiting professor at the Massachusetts Institute of Technology (MIT). Dr. Goodwin and Rosalind Picard developed the MIT course “Autism Theory and Technology,” with the support of an Autism Speaks research grant.
I’m delighted to take this opportunity to update you on our research developing and evaluating innovative technology for autism. Our goals are two-fold:
First, we seek to use unobtrusive monitoring to gain insights into the internal and behavioral challenges faced by those with autism – especially those who can’t express themselves verbally.
Second, we seek to deploy this technology in real-life settings where it can help children and adults with autism and their families.
This work involves developing, deploying and evaluating state-of-the art technology including video systems, microphones and wearable biosensors.
Importantly, we’re moving this technology out of the laboratory and into real-world settings where it’s needed most: clinics, classrooms and homes. Here is a sampling of what we’re doing:
Autism tech in the clinic
As families and autism clinicians know first-hand, diagnosing autism can be challenging. It’s also subjective, drawing on a high level of expertise that’s not available in every community. We hope technology can add a measure of objectivity to the process. We likewise hope that these systems can enable experts to make observations from distant locations.
We’re doing this by instrumenting all three participants in the autism evaluation: the clinician, parent and child. (See illustration above.) The instruments include wearable biosensors and audiovisual recording systems. These instruments are recording while the evaluator uses the diagnostic checklist known as the Autism Diagnostic Observation Schedule (ADOS).
As part of our research, we also have diagnostic specialists making notations of the relevant behaviors they observe during the evaluation. We collect all this information for later review and analysis.
Our aim is to better understand the complex behavioral and physiological interactions between the three participants of a child’s evaluation. I’m particularly hopeful that we can use this information to better understand a child’s strengths and weaknesses. At the same time, we’re looking for objective data that supports the evaluator’s and parent’s impressions.
If these models prove useful and reliable, the next step is to see if we can accurately analyze the information in automated ways. This might include machine learning and digital-signal processing. This could help reduce waiting time for evaluation and services.
Importantly, we have to ensure that this monitoring doesn’t disrupt the behavioral evaluation. So we continually explore ways to pare down the tech to a minimum.
I believe this line of research has the potential to advance how we objectively discriminate autism from other developmental issues. I think it can likewise help us understand and characterize autism’s great heterogeneity – or many “subtypes.”
These capabilities, in turn, can guide personalized interventions and advance understanding of underlying genetic and neurobehavioral signs of autism and related conditions.
Autism tech in the classroom
We’re also using time-synchronized video cameras and wristband biosensors in classroom settings. In this way we’re capturing objective information that helps us understand and prevent challenging behaviors such as self-injury, tantrums, wandering and aggression.
For instance, we want to know whether biosensors can help teachers and aides monitor their student’s internal arousal states. In this way, we can look at relationships with internal emotional states before, during, and after a problematic behavior has occurred. Ultimately, we want to help teachers and aides direct their attention to students who need help controlling their emotional state before they engage in a challenging behavior.
Autism tech in the home
We’re currently working on two initiatives that deploy technology in the home.
One – in collaboration with Cathy Lord, of Weill Cornell Medical College – starts with parents describing their children’s mood and behavior in daily journal entries. We combine this information with recordings from wristband biosensors and a voice-recognition system called LENA.
LENA records children’s voices and surrounding sounds, including other voices. This enables us to understand whether a vocalization is in response to another person or situation. LENA also allows us to analyze the pitch, volume and speed of the vocalizations. This can be useful in measuring a child’s mood, or what we call “affective state.”
By combining parent report, biosensors and LENA, we hope to produce a system that can help us monitor how well a given intervention or medication is working in the natural environment of the home.
Our second, related project uses commercially available computers, cameras, wristband biosensors, an FDA-approved sleep monitoring device and a smart phone app that allows parents to record and time-stamp snapshots and their behavioral observations.
In essence, we’re trying to give parents a ready set of tools to gather detailed information on their children’s development and response to treatments. In the process, parents can assist their child’s clinical support team and even collaborate in autism research.
At the same time, we’re learning how to introduce technology to families in ways that don’t intrude on their daily life and are respectful of their privacy.
Technology’s potential to transform autism treatment
Today, so much of autism intervention is trial and error. We need objective ways to help us personalize interventions for each child or adult. Technology can be an important part of this work.
We know we’re being very ambitious and still have lots to do. We’re in the “let’s build it; let’s try it” stage. At the same time, we’re collecting unprecedented amounts of behavioral information. So we have to develop new methods for analyzing massive amounts of data before we can use it to make new discoveries. Please stay tuned!
In closing, I want to extend a special thanks to the many families, teachers, and clinicians participating in this important work.
For more on Dr. Goodwin’s research, check out this video by Northeastern University:
Editor’s note: To learn more, see Dr. Goodwin’s 2008 discussion paper, “Enhancing and Accelerating the Pace of Autism Research and Treatment: The Promise of Developing Innovative Technology.” It helped open a new field of autism research.