Challenge Discussed : Artificial Intelligence seems to have a large-scale impact on radiology – but what really is the market opportunity, the size of the market, and what stands in the way of achieving adoption?

Speaker : Sid Shah

Key Takeaways :

·        AI is important for radiology today. It’s evident because there are more than 100 startups active in this space until 2018, and every major medical imaging company is also betting on AI. Radiologists too, have now moved beyond the question of whether AI will replace them. They know that’s not true. Instead the question being asked is, how can it help me, and how will it make my work easier?

·        As I discuss the economic ROI for the industry as a whole, I hope you will be able to view the entire space with a new lens.

·        Before I begin, a quick background – we forecasted the global market for AI in radiology as close to $1.5 bn by 2022. Another firm’s estimate pegs the global value at about $2 bn or so.. This value in billion dollars in just 3-4 years might raise eyebrows for you, and raise new questions, but let me propose a different way of looking at it.

·        Let’s take a step back, and look at the ENTIRE medical imaging space. The entire market of medical imaging services – including professional services, diagnostic and imaging equipment, imaging informatics and everything else accounts for more than $250 bn in 2022. Now as I build this story, hold that thought in your mind – the figure of $250+ bn in 2022.

·        The second part of this story is this: if we look at the total revenue for AI in radiology, and add up all annual revenues up to 2022 – the total comes to approximately $3 bn.

·        There are a couple of takeaways here. First, think of the $3 bn value, in view of the larger $250 bn value – it’s barely over 1% of the total industry size! So that value shouldn’t really raise concerns on a major disruption.. Sure there’s disruption – but in the good way of how a radiologist can actually utilize AI.

·        Now for the second takeaway. If the $3 bn cumulative is what the AI vendors will earn, it is the same amount of money that radiology organizations will spend. I hope that’s clear – I’m looking at the flip side – revenue for the vendor, is the spend for the provider. So let’s deem that $3bn value as an investment that providers will make. But what will be their ROI, from investing $3 bn in AI technologies? So we built some assumptions here – AI holds the promise of two kinds of savings for providers. One around productivity and efficiency, and another around diagnostic accuracy – I’m obviously over simplifying this for the sake of explanation. But based on this, we estimate that providers would, cumulatively, save north of $15 bn until 2022. That’s more than a 5x ROI for the providers…

·        Of course, we have challenges today on actually measuring those savings – primarily because we don’t use any kind of metrics to measure efficiencies to compare them post-AI implementation.. So how do you really measure those savings.. And this would be important in convincing the radiology organizations to invest that $3 bn in the first place.. A major problem statement withholding the growth of the industry is adoption – especially because there are questions around reimbursement by insurance companies, and then without that major factor, what business models for vendors will work, which ROI justifications for hospitals would work are all unknowns as yet.

·        This discussion is definitely not something that can be sorted out in one podcast, there’s much more to all of this. But at the end of the day, AI is being developed, it is being adopted. There’s more to it than just image analysis.

·        Ultimately there are growth opportunities for the industry. The most logical one that emerges is around automating and accelerating manual image analysis tasks.

·        And another one from the ten we have, is that of the AI development services – when its AI, you need data to train it. But that data must be extracted, curated, anonymized, annotated, and then build the AI algorithm and validate it, and the vendors who offer these services can benefit from this trend!

On that note, I hope you have enjoyed this session.

Please join us for future podcasts and become a member of our Leadership Council by emailing us at [email protected].

Thank you for your time, and Have a nice day!

Related Keywords :

Artificial intelligence, AI, augmented intelligence, ambient intelligence, anatomical intelligence, assistive intelligence, applied intelligence, deep learning, machine learning, computer vision, medical imaging, radiology, CT, MRI, XRay, ROI, savings, startups, use cases, revenue forecasts, market size, growth opportunities

Website : www.frost.com

Hosted on Acast. See acast.com/privacy for more information.