The genesis of an idea.
The Analytic for Hemodynamic Instability (AHI) and ultimately Fifth Eye originated from the idea that clinicians needed a better early warning indicator of an oncoming patient crash using existing physiologic information. The goal was to provide doctors and nurses an intuitive heads-up so they could take action to save the patient’s life.
In 2014 a team of physicians, nurses, data scientists and computer scientists began exploring a data set developed by the U.S. Army that simulated hypovolemia (hemorrhage) on healthy volunteers. Using sophisticated machine learning techniques, the team looked at a variety of patient data streams and ultimately settled on using a single real-time stream of rich, information-dense ECG data to develop a proof of concept analytic that outperformed traditional vital signs in predicting patient deterioration.
From army to academic hospital.
To further develop this concept, the University of Michigan developed a data collection framework, software tools and infrastructure to collect and store a vast dataset of deidentified patient waveform and numeric data. For four years, the Michigan Medicine team, that would eventually become the Fifth Eye team, collaborated with internal critical care experts to do a deep retrospective dive into cases to determine how early the AHI analytic could predict the collapse that led to a Rapid Response Team (RRT) call. Results from their work show AHI consistently and reliably predicts the impending hemodynamic instability minutes to hours and sometimes even days in advance.
The Fifth Eye team deliberately built our clinical analytic on a single real-time stream of very rich, very information-dense ECG data which is widely used on many different types of patients today. This means that many patients will have the streaming data needed to use AHI – and AHI does not require baseline data and generates a first prediction with just 120 heartbeats of data.
We have used customer-driven product design from the earliest stages. The core technology was developed with frequent and in-depth interactions with doctors and nurses to shape the direction of the technology development.
From academic hospital to the world.
In 2018, Fifth Eye licensed the technology from the University of Michigan in order to transform it into an easy to use and powerful tool for clinicians around the world. The Fifth Eye team is completing the product development, preparing for U.S. FDA clearance, and working with potential early adopting partners who are interested in being at the forefront of healthcare improvements.
The key value proposition is in accurately predicting problems before they happen to aid in the recognition and rescue of the deteriorating patient…this makes it a novel next generation monitoring and predicting tool.
We have a couple of teams dabbling in "early warning" models but they have not distilled it down to a single ECG lead!
The linkage of an early response team with this technology would have a very high likelihood of being associated with a significant survival benefit.
This is an amazing breakthrough! I can see so many applications within the hospital. I would like to get you connected to our innovation center so that we could be involved in its development.
Yes, this is a game-changer. The industry absolutely needs a better detection system for hemodynamic instability.
All I can say is wow! That is awesome.