As imaging AI becomes more prevalent in the healthcare market, there is a growing question around whether it should replace or augment elements of existing radiology workflow. Some providers are already promoting broad, cloud-based AI solutions that connect into a wide range of disparate systems and require wholesale replacement of large chunks of existing IT infrastructure in order to accommodate them. But how well does this kind of seismic change benefit the radiologist?
We have seen numerous examples over the years of technologies that were touted as being the harbingers of rapid and significant changes for the industry but, in reality, turned out to be much more considered evolutions over 10 to 15 years. The move to digitization through PACS was something that took a long time to evolve, and in fact, many of the structures and the companies are the same, even though the underlying technology has fundamentally changed. Another more recent example is advanced visualization, which people thought would change reading forever and in a very dramatic fashion, but most radiologists today still read axial slices. AI has utility all the way through from when the protocol is assigned at the modality, through to prioritization and timing on the worklist, through to viewing results in the viewer and reporting, and then downstream to the EMR. But, while the future of IT may well be AI-based, the existing IT infrastructure of a radiology department is typically interdependent and heavily customized, and replacing large parts of this in the short-term for the goal of getting AI is a significant undertaking. So, while AI is an incredibly useful and powerful tool in radiology, history tells us that it is likely to be more beneficial to radiologists, clinicians, and patients to augment existing systems than rip and replace them wholesale. Click below to learn more about the value of AI in Radiology in our newest White Paper.