Canon Medical's new diagnostic and sports centre in Sheffield is aimed in part at helping to expand the company's AI capabilities.

Canon Medical is betting big on a new diagnostics and sports facility in Sheffield, UK. The medical equipment manufacturer has invested £14m ($17m) in what will serve as the new home of basketball teams the Sheffield Sharks and Hatters, as well as housing some of the company’s top medical devices.


Canon Medical Arena, Sheffield. Photo: Canon Medical.

Canon claims the centre is the first of its kind, and sees the centre as a potential model for the future of sports health. For its part, it hopes to recoup the investment through gathering diagnostic data to train its medical AI, currently in development in Scotland.

Medical Device Network attended the launch and toured the medical facility, which is home to many Canon products including its Aquilion Prime SP CT scannerApilo i800 ultrasound machine and its Vantage Galan 3T high-powered MRI machine.

The suite clearly slants towards sports medicine, which is unsurprising given the setting. Diagnostics centre operator LivingCare estimates that, once up and running, the centre will be able to process around 5,000 patients a month, of which around 20% will be professional sportspeople.

The remainder of patients will come from LivingCare’s set of private patients, as well as the National Health Service (NHS) through public-private partnership. Despite claims that the centre will ease the backlog of NHS patients in Sheffield, the cost of this service to the public body is unclear. Investment Monitor has reported that the NHS spends around a quarter of its budget on private providers.

The benefits of integrating sports and diagnostics are twofold. For teams, players can be off the court and into a medical facility in under an hour after being injured. This means that teams and players can know what is wrong sooner, allowing for the creation of customised recovery regimens. The facility also has the capacity for minor surgery, allowing more serious injuries to be dealt with without having to go elsewhere.

AI diagnostics

From Canon’s point of view, the real benefit is data for its AI diagnostics tool. Though the project is still in its infancy, Canon hopes to go beyond simply diagnosing patients. Managing Director of Canon Medical Mark Hitchman told press at the event that he wanted data that could “not just diagnose, because that’s starting to emerge already, but do it safely. Or one step further, dare we think about predicting disease from other biomarkers in five, ten years? We dare.”

AI’s increasing use in the medical industry has been the cause for both celebration and concern. As its capacity to diagnose increases, it could cut costs and free up doctors to see more patients, which is vital given the limited resources many hospitals face, particularly in the UK. However, initial waves of optimism around AI in drug development have been stifled somewhat after many AI-designed drugs failed to pass clinical trials.

Canon is taking a moderated approach, choosing to hold off on bringing its product to market until it is sure of long-term accuracy. The company is developing a governance model that compares AI with another baseline AI and also separately with a doctor in real-time to ensure that accuracy does not drift, which is possible when AI ‘learns’ from other AI models.

When asked about market readiness, Hitchman said: “It’s like when you’re building a unique building. You never know when the ribbon cutting will be.” After a brief pause, he elaborated: “Two years. I don’t think we’ll ever really replace doctors, [but] I think we’ll help them be more efficient and reduce error rates a lot.”

This approach is likely to pay off for Canon, particularly given the quality of data it will be able to gather from elite sportspeople, giving a valuable insight into how the human body operates when pushed to its limits. Hitchman’s honesty is refreshing in a world where many have bought into the AI hype without considering the vital importance of safety and accuracy.

Carbon neutral?

The facility also claims to be carbon neutral, though this is achieved in part through the purchase of carbon credits, provided by C02balance. C02balance is a founding member of the International Carbon Reduction and Offset Alliance, a trade body for voluntary carbon offset providers and is accredited by the Gold Standard, an organisation founded by the WWF.

The voluntary carbon credit market has been criticised for predominantly operating based on counterfactual carbon reductions. The amount of carbon offset is often calculated assuming that, for instance, patches of forest purchased by carbon credit providers would have been cut down were they not bought by the company. A Guardian investigation early this year found that over 90% of rainforest carbon offsets offered by the largest certifier were worthless.

Gold Standard holds itself to rigorous standards and has recently been found to have strong methodology in an independent review. However, the standard uses suppressed demand in its calculation of carbon credits, meaning that companies such as CO2balance are able to claim carbon reductions that do not currently exist.

One project that Canon Medical bought from reduces carbon by repairing boreholes in Uganda, allowing citizens access to clean, fresh water. This reduces carbon output by preventing people from needing to boil water to purify it for drinking. However, Medical Device Network has found that only 16.8% of the 45,000 people affected by the project currently boil water for purification purposes. While almost all of the remainder have said that they would if they had the fuel, making this calculation based on suppressed demand allows CO2balance to assume savings of 60,000 tons of greenhouse gas emissions per year. This is equivalent to the 2021 net emissions of 40,000 Ugandans, per Our World in Data.

In conversations with Canon Medical and Medical Device Network, CO2balance stressed its commitment to methodology, noting the “rigour and process that we put our projects through”.