How imaging practices can evaluate which AI products will drive clinical and business benefits for their practice.

Artificial intelligence and machine learning technologies are sweeping the globe and delivering deeper insights and radical efficiencies across multiple industries. The healthcare industry and, in particular, the medical imaging sector, have been slower to adopt AI than other industries. But why?

What’s holding AI back in imaging?

Well, for starters, healthcare is a highly regulated environment. Products being used in a clinical setting need to have regulatory clearance, and it’s only in the last couple of years that a significant number of AI solutions have been approved for clinical use.

Then there is the complexity of healthcare IT infrastructure. It’s not easy to deploy clinical AI applications, especially when they have to be integrated into all elements of workflow to make them truly useful. You can’t just throw an AI result to a radiologist and hope they can create value – it has to appear where they need it and when they need it.

And don’t forget that AI is still relatively immature, and the market is very noisy. It can be difficult for anyone, but especially busy clinicians, to evaluate the available solutions and work out which ones will work across their patients, radiologists, referrers and modalities. In the absence of widespread reimbursement, there is a pressing need to building the business case for AI in order to allow radiology practices to determine how they make a return on the procurement of AI clinical applications.

Challenges still exist

Even with more regulatory clearances in place, and platforms like Blackford helping to make multiple AI application deployments more manageable, it can still be a challenge to evaluate the clinical and revenue performance of products in real life. For many forms of AI, although the medical image and radiologist are at the centre of detecting or characterising a condition, the real value is unlocked downstream – in oncology, neurology, surgery and so on. The biggest challenge for AI is to link the downstream value to radiology cost.

AI holds great promise for improved patient outcomes, and its potential for broader healthcare economic benefit is huge. But the fragmented nature of the market means that, unless a sensible commercial business case can be made, adoption of AI solutions will be limited. Part of solving this challenge is having the ability to continuously measure impact of AI solutions, a key element of the Blackford Platform.

At Blackford, we have developed a platform that integrates across the whole of the radiology workflow, providing access to a broad portfolio of applications that address the majority of the clinical use cases tackled by AI. The absolute focus for us – and all of our customers – is using our platform to build and prove clinical and business models for adoption of these AI solutions in healthcare.