TORONTO, Oct. 12, 2026 | A team at University Health Network (UHN) in Toronto is using AI to turn flat medical scans into interactive 3D models of a patient’s tumour and surrounding anatomy, with the aim of helping surgeons plan operations and helping patients understand them, Global News reports. The technology has not yet completed clinical trials.

The researchers call the result a “digital twin.” Wagner Souza, a staff scientist at UHN and assistant professor at the University of Toronto, described it as “you cloned in 3D.” The model can be viewed from different angles to show depth, the lesion’s location and how the area could look after removal. Global News’ examples involve breast cancer, including a lumpectomy and sentinel node removal.

Amin Madani, an endocrine and acute care surgeon at UHN, told Global News that surgeons currently review 2D MRI and CT scans and memorize the anatomy before operating. A 3D model would let them plan and set expectations ahead of surgery. Madani sees AI-driven guidance during operations as “the next frontier” in surgery.

For patients, the model is a communication tool. Adina Isenberg, a breast cancer patient diagnosed nine years ago who has joined the team to explain the technology to other patients, said seeing their own model could reduce unnecessary medical trauma. Souza said it lets patients take part in decisions: “It’s empowering, because you know where you’re fighting.”

The team has not published trial dates, sites, sample sizes, accuracy figures or outcome data, and Global News does not report funding or regulatory approval. Early use suggests possible benefits in the operating room, but that is not the same as clinical evidence.

There are also technical gaps. Patients are scanned lying face down but operated on face up, which changes the position of breast tissue. The team is exploring whether to model the physical properties of tissue to account for the shift. “There are some things we can still not do, but we’re working towards being able to achieve” those goals, Souza said. Isenberg noted the model cannot predict how a patient’s body will respond to treatment.

How is UHN using AI to model tumours?
UHN researchers use AI to build a 3D “digital twin” of a patient’s tumour and surrounding anatomy from 2D scans. Surgeons can view it from different angles to plan an operation, and patients can see what a procedure involves. It has not yet completed clinical trials.

Is the AI tumour model ready for patients?
Not yet. The technology still has to go through clinical trials, and the team has not reported trial details or outcome data. Researchers are also working on how scan position differs from surgical position.