RADiCAIT (Techstars 2024) is a medical technology company that is transforming radiology by turning standard CT scans into high-fidelity, in silico PET scans. By leveraging deep learning, the relationship between structure and function, between anatomy and physiology, is decoded — delivering critical clinical insights from existing routine scan data, the company is making the precision of PET imaging as fast, affordable, and accessible as a common CT.

Emilie Vallauri, Global Program Manager at Techstars, sat down with Dr. Seán Walsh, CEO and co-founder, to discuss how his team of "surgeons and scientists" is working to make PET-level insight the new standard of care.

Emilie Vallauri: To kick things off, can you introduce yourself and your background?

Seán Walsh: I’m a medical physicist by training. I studied physics in Ireland and became clinically qualified to work in both Europe and the US, which is a bit of a unique "feather in my cap". However, I turned down clinical work to become an innovator. I spent years in research — specifically in statistical modeling and medical imaging — and published a paper on radiomics that became a very impactful piece of research in its field, cited more than 6,000 times in other academic works over the last 20 years. Before RADiCAIT, I co-founded another medical imaging startup that scaled to 50 employees and worked closely with big pharma.

Emilie: Who are the other members of your founding team, and how did you meet?

Seán: I first met my co-founder, Regent Lee, while I was doing an MBA at Cambridge. He’s a Professor of Interdisciplinary Innovations at the University of Oxford and a surgeon-scientist who originally invented the technology. 

The team also includes our CTO, Dr. Sina Shahandeh, a computational physicist who went through a major hyperscaling journey at a company called Ecobee, and James Fox, our COO, who is a retired US Naval intelligence officer with a background in high-pressure operations and image analysis. They joined both Regent and I to complement our existing set of expertises.

Emilie: How did the idea for RADiCAIT start? What problem were you solving?

Seán: It started over lunch with Regent. He told me he had found a way to convert a CT scan into a PET scan using AI. At first, people were skeptical because textbooks say they are completely different. But I had seen evidence in my previous work that machine vision models could pull information out of CT scans that the human eye simply couldn't see.

The problem is that while CT scans are everywhere, they only show anatomy — like a map. PET scans reveal disease activity, which is much more valuable, but they are scarce, slow, invasive, and expensive. We realized if we could produce a PET scan from a CT, we could provide PET-level insight at the scale of CT.

Emilie: For clarity, what does RADiCAIT do and why is it so important?

Seán: We take a CT scan — which every hospital has — and process it to produce an equivalent in silico PET scan. It takes about two minutes instead of the month-long wait often associated with traditional chemical PET scans. This is vital for oncologists who need immediate answers to make life-saving decisions for cancer patients. It removes the need for radioactive injections and specialized nuclear medicine departments for every single scan.

Emilie: Who is the paying customer, and how does the business model work?

Seán: The oncologist is the primary persona who needs the answer, and they work with the radiologist to order the scan. They are our “users,” so to say. However, the "purchaser" is the hospital procurement officer. We are both a revenue driver and a cost reducer for them. Our solution is significantly cheaper than a chemical PET scan. Furthermore, we unlock a massive market for smaller clinics that have CT scanners but don't have the multi-million dollar equipment required for traditional PET imaging.

Emilie: What sets RADiCAIT apart from competitors? 

Seán: We don't have a direct competitor. The FDA even told us that nothing like this has been cleared yet. We have three granted patents in Europe and America covering both our methods and applications, which gives us strong defensibility. Because we are the first to do this, anyone following us would have to design a new technology from scratch and run their own years-long clinical trials. We have a huge first-mover advantage.

Emilie: It’s been quite a busy year for RADiCAIT. What’s your proudest recent milestone?

Seán: Our clinical pilots. We moved from Oxford to the US and, within our first year, signed up four major health systems: Mass General Brigham (MGB), UCSF, Emory, and Hamilton Health. 

Demonstrating that our technology works in these top-tier environments and getting validation from US payers was a huge moment for us. We also recently closed a $1.7M funding round from investors like Frontline Ventures, Gurtin Ventures, and of course, Techstars.

Emilie: You are Irish and met most of your co-founders in Europe. Why did you decide to build RADiCAIT as an American company?

Seán: If you are building a transformational company that wants to truly disrupt healthcare, the US is the best place to do it. US VCs are more likely to take a "big swing" at a project like this, and the US healthcare market is the most pioneering in the world. We are a Delaware C Corp and are focused on working with the best health systems in America.

James Fox, our COO, is located in the United States and focusing on developing that market for RADiCAIT. I’m based in between Europe (Ireland, the Netherlands, & UK)  and the United States, spending half of my time on the ground with our partners.

Emilie: What are the next steps for RADiCAIT?

Seán: We are currently turning ‘what if’ into ‘when’ — so the next chapter of our story isn’t just delivering an important milestone; it is a fundamental re-imagining of the standard of care in radiology. We’re building a future where no critical image-based decision in Oncology, Neurology, and Cardiology comes too late and no patient is left behind. We’re raising a seed round to fund our formal clinical trial, which is the final step for FDA approval. We expect to have our first patient processed and reimbursed by mid-2027. From there, it’s all about scaling as fast as possible to reach the 100 million people diagnosed with these diseases every year.

Read More