This is an image of Dr. Rosalind Picard presenting at INSAR. She is standing at a podium in a large auditorium. Behind her, a screen has an image of a person wearing what appears to be a watch, but without any numbers on it.

New Wearable Detects Meltdowns Before They Happen

A new wristband may give autistic people and our caregivers warning that a meltdown is coming up to three minutes before they happen. A version of the wristband is already FDA approved and on the market to help detect seizures. Dr. Rosalind Picard of MIT presented the results on Friday at the 2018 International Society for Autism Research annual meeting in Rotterdam, the Netherlands. ­­

During her keynote, Dr. Picard emphasized the importance of looking beyond neurotypical facial expressions and cues to determine how autistic people feel. Dr. Picard began her keynote by stressing the importance of autistic feedback in her work. She recalled an autistic friend’s observation that “my biggest problem is not understanding the emotions of others. My biggest problem is you are not understanding my emotions.”

Alexithymia, or difficulty understanding how one feels, is a common problem in the autistic community. This can often lead to situations in which someone does not notice that they are on the verge of meltdown until they have reached the point of no return. This can be embarrassing and frustrating for many autistic people. Seemingly unpredictable meltdowns can also be a concern for caregivers, as they can significantly limit their children’s community participation.  Dr. Picard’s new breakthrough may help autistic people and our caregivers head off meltdowns before they start, or at least get a head start on helping during crisis.

The device is made by a Massachusetts-based company called Empatica. It detects a person’s heart rate, sweat levels, electrical activity, and skin temperature. The sensor can predict an impending meltdown with up to 70% accuracy. When the device is further adjusted to better reflect individual differences, the accuracy rises to 84 percent.

Matthew Spiegel, one of the lead co-investigators told Spectrum News, “If we use some objective signal to predict when someone is going to have a [meltdown], we don’t have to infer anything, and we can get ahead of it, which is a real shift.”

Researchers worked with 20 nonspeaking autistics  who had been admitted to an inpatient psychiatric unit. The participants wore the bands on their ankles as they went about their typical day. Wristbands were not well-tolerated by the people testing the devices.

The research team collected data for 87 hours, during which the participants had 548 meltdowns and instances of aggression. They looked at variables that might predict a meltdown from three to one minutes before an episode. Then, they used a machine-learning algorithm to analyze their data.

Among other discoveries, researchers found quantitative data indicating that stimming helps calm autistic people who may be experiencing elevating stress levels. Dr. Picard noted, “sometimes our bodies know how to calm us.”

Researchers found that the device works best when it includes all of the measures that the device detects, and when it is personalized to each individual person.

A larger study is necessary to determine how useful the device might be to a wider population of autistic people, according to Dr. Goodwin.

The device is not yet ready for clinical use, but many self-advocates and family members at INSAR were excited about potential applications. Shannon Rosa, a parent advocate from Thinking Person’s Guide to Autism told NOS Magazine, “I first got excited about wearable tech for autistic self-regulation possibilities after Matthew Goodwin’s presentation at the 2012 [INSAR] meeting in Toronto, and was so glad Picard demonstrated real progress in making it so.”

On social media, Dr. Sue Fletcher-Watson, a researcher from University of Edinburgh, noted that privacy concerns were under-addressed. In particular, she expressed concern for the risks of 24/7 data collection and the difficulty of getting truly informed consent from people with complex communication needs. “This line of research could do with a good ethical framework to help us consider when and how and why it’s OK to gather intimate data from people who can’t give traditional informed consent.”

In this reporter’s own conversation with Dr. Rosalind Picard, one issue that emerged was the lack of user interface for her device. Currently, the data goes to a computer that a caregiver has access to. In terms of improving self-regulation, an easy-to-understand user interface is necessary. Dr. Picard agreed that this was an important area of development before her product moves further through the FDA approval process.

 

 

4 thoughts on “New Wearable Detects Meltdowns Before They Happen”

  1. Most intriguing as I am desperate to know the inner workings of my autistic husband’s experience of life which he is unable to communicate to me. If even one unpleasant episode could be forecasted and avoided each day, it would surely raise the quality of our lives together. Thank you for this promising work and information. It signifies hope versus exasperation and failure and nothingness.

    1. Hi Valerie,

      I’m the editor of this website. I’m also autistic and in a long term relationship with someone who isn’t. If your husband is hurting you, that’s not autism. That’s him. No wearable is going to fix the issue. I’m so sorry that you’re in this situation, and strongly encourage you to find someone else who will treat you with the care and respect you deserve, autistic or not.

  2. I believe that this technology would greatly improve/aide persons w autism. My 8 y o grandson has autism & I struggle w pain watching him become frustrated & unable 2 express himself. He is nonverbal but tries sooooo hard 2 repeat what I teach him. Whenever this gets fda approval, (if affordable) I will be 1st in line! I think this would really really help alot of people, autistic & non-autistic! Even a 3 min. warning could seriously help 2 try 2 jump-start meltdown reversal. Ty

  3. We have an Embrace. It’s one of the original seizure detection models.. I have to tell you that it doesn’t work. It’s not connecting to or notifying caregivers. After many attempts to contact the support staff we were told that we must be in the same room to be notified, or it’s appearing to work from their end. Having notifications for heart rate, pulse and temperature fluctuations would be wonderful.. my child is also autistic. But so far the model we have was an expensive mistake.

Leave a Reply

Your email address will not be published. Required fields are marked *