People at risk are always vigilant for the signs of an ongoing stroke, but no one can be vigilant when they are asleep, which means that thousands of people experience a “wake-up stroke” that is not recognized until hours later. Time medicals Brain monitoring wearables could help sound the alarm and get people to the hospital fast enough to mitigate the damage from the stroke and potentially save lives.
A few decades ago there was not much that could be done to help a stroke patient. However, an effective drug came into use in the 1990s, and surgery was introduced a little later – but both must be given within hours of the onset of the stroke.
Orestis Vardoulis and Urs Naber founded Zeit (“Zeit”) after seeing how resources were being used to reduce the delay between receiving an emergency call about a stroke and receiving the necessary therapy from the victim. The company is part of Y Combinator’s 2021 summer cohort.
“In the past you couldn’t do anything, but suddenly it was really important how quickly you got to the hospital,” says Naber. “As soon as the stroke hits you, your brain begins to die, so time is the most important thing. People have spent millions cutting the time between calling 911 and being transported and from the hospital door to treatment. but no one speaks of those hours that are happening before the emergency call – so we realized that we have to be innovative here. “
If only the stroke could be detected before the person even knows it is happening, they and others could be alerted and taken to the hospital long before an ambulance would normally be called. As it turns out, there is another situation where this needs to be done: in the operating room.
To illustrate this, an EEG signal that changes its character can be quickly recognized by the algorithm.
Surgeons and nurses performing operations apparently closely monitor the patient’s vital signs and have learned to recognize the signs of an impending stroke through the EEG monitoring their brain waves.
“There are certain patterns that people are trained to capture with their eyes. We learned from the best neurologists how to visually process this data and we developed a tool to automatically detect this, ”said Vardoulis. “This clinical experience really helped because it helped define features within the signal that helped us speed up the process of deciding what is important and what is not.”
The team developed a soft, wearable headband with a compact built-in EEG that monitors relevant signals from the brain. This data is sent to a smartphone app for analysis by a machine learning model trained on the patterns above, and if something is detected an alert is sent to the user and pre-defined caregivers. It can also be set to automatically call 911.
“Most of the data we analyze comes from the operating room,” says Vardoulis, where it can be compared immediately with the ground truth. “We saw that we have an algorithm that can reliably record the onset of events in the OR without false alarms.”
That should translate well to the home, they say, where there are actually less complicating variables. To test this, they are working with a group of high-risk patients who have already had a stroke; the months immediately following a stroke or related event (there are several clinically differentiated categories) are dangerous when second events occur frequently.
Photo credits: Time
“Right now we have a research kit that we ship to people involved in our studies that includes a headband and a phone. Users wear it every night, ”said Vardoulis. “We are preparing for a path that will allow us to go commercial sometime in 2023. We’re working with the FDA to define the clinical evidence needed to make this clear. “
They have received a Breakthrough Device classification that enables them (like stroke rehab company BrainQ) to move forward quickly with testing and certification.
“We will start in the USA, but see demand worldwide,” said Naber. “There are countries where aging is even more prevalent and the support structure for disabled care is even lower.” significantly lower.
The plan for now is to continue collecting data and partners until they can do a large-scale study that will almost certainly be required to get the device reimbursed (i.e. covered by insurance) from direct to consumer. And while they currently focus entirely on strokes, the method could be adapted to check for other neurological disorders.
“We hope for a future in which anyone at risk of stroke will have this device issued,” said Vardoulis. “We really see this as the missing piece of the puzzle in the stroke care continuum.”
Source link