This AI tool can predict organ failure 48 hours in advance

It was right after his grandfather died of multi-organ failure, five days after being admitted to hospital for a heart attack, that PhD student Samyakh Tukra started wondering if there was anything he could do to improve outcomes in intensive care units (ICUs).

“I found out that there are more than 85 million patients admitted to the ICU annually for potential organ failure, yet only up to 50% of them survive,” Tukra says. “It was no longer just about my grandad.”

As Tukra started researching how clinical decisions in ICUs happen, he realised that doctors have to take into account multiple sets of data from different sources, both structured — such as live vitals from patient bedside monitors — and unstructured — such as patient notes, radiology reports, and more.

Soon after, Tukra launched Third Eye Intelligence, an AI-powered solution that helps clinicians intervene early thanks to an automatic machine learning algorithm capable of processing both structured and unstructured data.

The tool can predict the onset of organ failure 48 hours in advance, with 91% accuracy. It also reduces false positives, improving alarm fatigue for clinicians and freeing up already scarce resources in the ICU.

Third Eye Intelligence has already teamed up with three NHS trusts — a total of eight hospitals — and is currently fundraising to conduct initial clinical trials. The startup has also partnered with the National Institute for Health Research to measure how these early warnings can improve quality of life for patients.

Tukra joined Conception X on the final year of his PhD in AI and 3D Computer Vision at Imperial College London.

“Although my startup is in predictive healthcare analytics rather than 3D reconstruction and computer vision, it goes to show the beauty of the domain I work in,” he says. “AI is like a golden bullet that can bring a solution no matter what problem you’re looking at, as long as there’s a data source. I could apply what I learned as part of my PhD to Third Eye, and vice versa, I was able to bring back to my PhD what I learned with Third Eye to improve some of my own decisions there.”

For Tukra, the Conception X programme has been a time of growth — both personal and professional.

“Before joining, Third Eye was more of an engineering project — I was focused on developing a certain tool and proving it works,” he says. “I feel like the company has grown commercially — I’ve been able to get the right commercial partners to move this idea forward — and through some of the meetings with mentors I have learned how to think like a founder, how to deal with conflict, how to organise my time and decision-making, and more.”

Tukra now plans to validate the product in a real-world scenario over the course of 14 months. The goal is to eventually apply for regulatory approval in the UK and the US by the end of 2024.

Asked if he has any tips for the new cohort, he says: “Figure out what you want out of the programme, so that you can reach out to the right experts in the network and maximise your time there. Know where you’re headed.”

--

--

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store
Conception X

Conception X

Venture builders creating deeptech startups from leading research labs and PhD programmes around the UK. conceptionx.org