This piece is part of a series highlighting local business growth in the Ann Arbor area. It is supported by Ann Arbor SPARK.
Readings from a single ECG sensor might soon be able to predict when a patient is going to "crash," thanks to technology developed by the University of Michigan’s Integrative Research in Critical Care group and spun out into a new company, Ann Arbor-based Fifth Eye Inc.
In late July the health predictive analytics company raised $2.36 million from Invest Michigan, the Biosciences Research and Commercialization Center, and 33 angel investors, allowing Fifth Eye's five-person team to turn the technology into a viable product.
Fifth Eye CEO Jen Baird is a serial entrepreneur whose resume includes serving as co-founder and CEO of Ann Arbor-based Accuri Cytometers and later as CEO of Pittsfield Township-based Accio Energy.
Baird says that when you watch a medical drama on TV and heard medical staff call "code blue," they are dramatizing something that really happens in real life. A patient may crash due to undetected bleeding or sepsis, but hospitals can't easily detect those risks until it's too late.
Today, the standard way to watch for a patient about to crash is to monitor their vital signs, looking for indicators of unstable blood flow through the patient's heart.
"But the body is so good at compensating for trouble, that vital sign being off is a late indicator," Baird says. Fifth Eye's technology predicts trouble in early stages by "picking up signs that the heart is working harder to compensate and keep the vitals normal as long as possible."
Baird says the tech was developed using "supervised machine learning." Patient data is marked by experts, and then computers analyze the notated patient information to pick out the difference in ECG signal between a patient whose blood flow is unstable versus one who is stable.
"It's super cool," Baird says. "They developed new math and some fancy signal processing to take an ECG signal that is not a clean signal in really sick people, and figure out what to use for streaming analysis and (the ability) to detect a problem in near-real time."
The same technology, which creates a "score" for the patient every two minutes, can tell a doctor quickly if an intervention is having a positive effect.
"In addition to predicting something bad, it can give feedback if something the doctor is doing is working," Baird says. "Today, a doctor might try something, and he won't know for a day or two if it's working. With this technology, they might be able to tell in 15 or 20 minutes."
The recent funding round will allow Fifth Eye to take its underlying technology and convert it into a software product. Fifth Eye is working with developers from Ann Arbor's Menlo Innovations to put the analytics into a product. Next, it will be submitted to the FDA for clearance so healthcare centers can start using it.
Baird says she expects a commercial product to become available to hospitals in the second half of 2019.
Sarah Rigg is a freelance writer and editor in southeast Michigan. You may reach her at firstname.lastname@example.org.
Photos courtesy of Michigan Photography.