Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
If you’ve ever had a PET scan, you know it’s an ordeal. Scans help doctors detect cancer and track its spread, but the process itself is a logistical nightmare for patients.
It starts with fasting for four to six hours before admission to the hospital — good luck if you live in the countryside and your local hospital doesn’t have a positron emission tomography (PET) scanner. When you arrive at the hospital, you are injected with a radioactive substance, and then you must wait an hour for it to be washed into your body. Next, you go into a positron emission tomography (PET) scanner and have to try to lie still for 30 minutes while the radiologists get the image. Next, you should physically stay away from the elderly, young people, and pregnant women for up to 12 hours because you are literally almost radioactive.
Another bottleneck? PET scanners are concentrated in major cities, because their radiation traces must be produced in nearby cyclotrons — compact nuclear machines — and used within hours, limiting access to rural and regional hospitals.
But what if you could use AI to convert CT scans, which are easily accessible and affordable, into PET scans? This is the playground Radikitean Oxford University spin-off, emerged from obscurity this month with $1.7 million in pre-seed funding. The Boston-based startup, which made it to the list of top 20 finalists in The emerging battlefield in TechCrunch disabled 2025just opened a $5 million raise to advance its clinical trials.
“What we’re really doing is we’ve taken the most restrictive, complex and expensive medical imaging solution in radiology, and replaced it with what is much easier, simpler and more affordable, which is CT,” Sean Walsh, CEO of RADiCAIT, told TechCrunch.
RADiCAIT’s secret sauce is its core model – a generative deep neural network invented in 2021 at the University of Oxford by a team led by the startup’s co-founder and chief medical information officer, Regent Lee.
The model learns by comparing CT and PET scans, mapping them, and picking out patterns in how they relate to each other. Sina Shahande, chief technologist at RADiCAIT, describes it as connecting “distinct physical phenomena” by translating anatomical structure into physiological function. The model is then instructed to pay additional attention to certain features or aspects of the scans, such as certain types of tissue or abnormalities. This focused learning is repeated many times with many different examples, so that the model can identify clinically important patterns.
TechCrunch event
San Francisco
|
October 27-29, 2025
The final image that is sent to doctors for review is created by combining multiple models working together. Shahandeh compares this approach to Google DeepMind’s AlphaFold system, an AI that has revolutionized protein structure prediction: Both systems learn to translate one type of biological information into another.
Walsh claims that the RADiCAIT team can mathematically prove that their synthetic or generated PET images are statistically similar to real chemical PET scans.
“This is what our experiments show, where the same decision is made when a doctor, a radiologist or an oncologist is given chemical PET or[artificial intelligence-generated PET],” he said.
RADiCAIT does not promise to replace the need for positron emission tomography (PET) scans in certain treatment settings, such as radiation therapy, which kills cancer cells. But for diagnostic, staging and monitoring purposes, RADiCAIT may make PET scans obsolete.
“Because this system is constrained, there is not enough supply to meet the demand for diagnostics and therapeutic diagnostics,” Walsh said, referring to the medical approach that combines diagnostic imaging (i.e., positron emission tomography) with targeted therapy to treat diseases (i.e., cancer). “So what we’re looking to do is accommodate that demand on the diagnostic side. The PET scanners themselves will have to make up for the shortfall on the therapeutic side.”
RADiCAIT has already begun clinical trials specifically for lung cancer testing with major health systems such as Mass General Brigham and UCSF Health. The startup is now seeking an FDA clinical trial — a more expensive and involved process that led RADiCAIT’s $5 million seed round. Once approved, the next step is to conduct commercial pilot trials and demonstrate the commercial viability of the product. RADiCAIT also wants to run the same process — clinical pilots, clinical trials, and commercial pilots — for colorectal and lymphoma use cases.
RADiCAIT’s approach of using AI to gain valid insights without the burdens of difficult and expensive testing is “broadly applicable,” Shahande said.
“We are exploring expansions into the field of radiology,” Shahandeh added. “We expect to see similar innovations linking fields from materials science to biology, chemistry and physics wherever nature’s hidden relationships can be learned.”
If you want to hear more about RADiCAIT Join us at Disrupt, October 27-29 in San Francisco. Learn more here.
