By Pragati Verma, Contributor
Two years ago, 25 year-old Morgan McGrath, suffered an epileptic seizure that left her unresponsive, frothing at the mouth, and not breathing. Embrace, a medical grade square-faced watch on her wrist, detected the seizure and immediately sent out an alert to her mother, who called the paramedics.
“I was in the hospital for a total of four days with three of those days being in ICU,” McGrath explained in a blog post for Empatica, the startup behind Embrace. “I believe the Embrace saved my life. It helps give peace of mind to me, as well as my friends and family.”
Embrace, the device McGrath wore, is the first smartwatch to be approved by the U.S. Food and Drug Administration (FDA) for neurology. It uses machine learning to monitor one of the most dangerous kinds of seizures, known as “grand mal” or “generalized tonic-clonic” seizures, and sends an alert to summon caregivers’ help.
Today, other smartwatch makers are beginning to monitor biomarkers — from heart rate, glucose, and oxygen levels to toxins and other physiological signs — to detect and predict serious medical episodes.
Inside the World of Biomarkers
Empatica’s journey began long before it ventured into medical features. Its core technology traces back to 2007, when a team at MIT Media Lab developed a wearable that measured changes on the surface of the skin that could communicate main components of stress, such as electrodermal activity and temperature.
Professor Rosalind Picard led the research at MIT Media Lab, and today she also works as Empatica’s chief scientist. The team at MIT Media Lab, according to Picard, “worked for years building wearable stress and emotion sensors, and then accidentally discovered we could pick up changes in the skin elicited by brain activity related to the most dangerous kinds of seizures.”
The project began as a series of studies at several epilepsy centers, certified as Level 4 by the National Association of Epilepsy Centers (NAEC), meaning that they provide medical and surgical evaluation and treatment for the most complex cases of epilepsy and seizure disorders. These studies contributed to the development of Empatica’s seizure detection algorithm that uses artificial intelligence (AI) to recognize the complex physiological patterns that are most likely to accompany a convulsive seizure.
After raising over $780,000 in a crowdfunding campaign in late 2014, the research project soon turned into a wearable device that looks similar to existing smartwatches on the market and, that same year, Picard, Matteo Lai, Maurizio Garbarino, and Simone Tognetti co-founded Empatica.
Passing the Test
Embrace, the company’s first major biomarker venture, works by measuring three kinds of data.
First, it detects the electrodermal activity that represents electrical changes on the surface of the skin. For most people, increased neural activity and stress elevate the sweat level on skin. Even if the sweat is too little to be seen or felt on the surface, pores below fill and increase electrical conductance enough to be detected by Embrace’s sensors. Embedded accelerometers and gyrometers then measure movement to tell the device if the person is likely to have fallen or is making sudden movements. Third, a thermometer picks up longer-term changes in skin temperature.
When the Embrace wristband determines that a seizure is happening, it relays this message to the user’s paired phone and sends an alert to all the individuals listed as caregivers in the Empatica Alert app.
Picard and her team tested Embrace on 135 patients diagnosed with epilepsy. Subjects were monitored for more than 6,500 hours over 272 days in a multi-site clinical study published in 2017. In the end, the device correctly identified every single seizure.
The patients in this most recent study were continuously monitored with video EEG, while also wearing an Empatica device. This helped researchers compare seizures identified by Embrace’s technology with seizures clinically labeled by independent epileptogolists, who examined the video-EEG data without seeing any figures used by Embrace.
The clinical validation was a big breakthrough for Empatica. However, co-founder and CEO Matteo Lai still wanted to take the product one step beyond being an effective medical device. His ambition was to create “the first wearable that marries world-class design with scientific-quality data.”
Relying on Data and Design
According to Lai, medical devices typically face a huge problem: While people might believe in their health benefits, they still don’t want to wear them. “They’re usually too bulky and uncomfortable,” he went on, “and people simply don’t want to wear them.”
Empatica is not the only company trying to fit a doctor on our wrists, cloaked in a sleek design. Alongside Embrace, iBeat HeartWatch monitors users’ heart rates, pulse strength, temperature, blood flow, and oxygen levels to look for signals that indicate a life-threatening cardiac episode. If the user displays signs of physiological stress, the watch will inquire if they are okay; with no response, the operating system will alert emergency authorities.
Despite other biomarkers emerging onto the scene, Lai is still confident in Embrace’s data-driven approach, as well as the company’s commitment to design. “A device that could win a design award, while being used as a lifesaving product is one of the keys to its success and an interesting lesson for healthcare,” he explained.
For now, Lai’s mantra remains clear: “Cutting edge technology and good design need to go together.”