Artificial intelligence in medical imaging
There was heightened buzz around the topic of AI at the 24th European Congress of Radiology, with many vendors citing the integration of data through AI as the next big trend – but what’s new? AI has been discussed at the show for several years, and each year there is more and more uncertainty around its integration into medical devices. That said, imaging vendors seem to be taking a bigger grasp of reality and appreciate the vision of improved imaging through enhanced and advanced data usage. Market leaders in radiology have incorporated some elements of AI into their product offering and claim that things are only likely to get better.
AI: What is it? What is it being used for?
The term AI is used for various forms of data usage.
The use of AI in imaging modalities will assist radiologists to help them in their diagnosis. AI will aid physicians at the point of care by collaborating data, sorting it, and analyzing relevant information in a digestible format. This will not only speed up diagnosis time but will help to improve confidence in “first time right” diagnosis and reduce the need for rescans. With advances in data collection throughout the care continuum, there is a wealth of information available to help assist and generate informative diagnosis, through the pattern of recognition. AI will further enable the utilization of vast data lakes to assess patients against population health statistics and further relating this back to the patient, to help improve detection of disease.
What is the next step?
As with all things new, there has been initial push back from some radiologists for the fear that the new technology will hinder rather than help them. Skepticism is further heightened by radiologists often complaining of feeling burned out, being over-loaded with new workflows to understand; digitalization is thought to be a heavy contributor in this. As more is being learnt about the capabilities of AI, and the realization that it will not replace the physician in the diagnosis process, interest is starting to peak. Further product developments are being made, as we learn more about the true capabilities of AI in imaging, and start to benefit from its inclusion in the care continuum. Vendors are pushing their solutions forward by demonstrating their role in improving work flow by increasing accuracy and efficiency, to subsequently improve patient outcomes. More than 50% of global healthcare leaders expect an expanding role of AI in monitoring and diagnosis.
Who is leading the market?
ECR saw a pavilion dedicated to artificial intelligence providers to the medical imaging market, enabling AI providers to demonstrate their capabilities. Exhibitors showcasing their niche algorithms included EnvoyAI, Aidence, Aidoc, Cure Metrix, and iCAD. In addition to niche-AI providers, medical imaging and IT providers also shared how they are integrating artificial intelligence into their solutions. The most notable vendors included Philips Healthcare, Siemens Healthineers, Carestream, GE Healthcare, and Agfa. Philips Healthcare launched the new Ingenia 3.0T MR system, which incorporates AI into the updated technology to improve patient set up and reduce scanning time. Carestream demonstrated its advanced imaging analytics software tools as part of its Clinical Collaboration Platform. GE Healthcare launched its Logiq E10 ultrasound scanner, which incorporates AI, claiming 10 times the processing power of previous systems.
What are the cost efficiencies?
As healthcare spending becomes increasingly scrutinized, providers are looking at the total cost of ownership in the review of long-term costs. Both healthcare providers and manufacturers are beginning to work in partnership to remove inefficiencies and provide more cost-effective solutions. Providers want the “first time right diagnosis,” subsequently reducing admissions and overall costs of healthcare provision. By incorporating AI into solutions, providers will be able to see significant cost savings over the lifetime of the product. This is further supplemented through product developments that improve the overall patient experience. Focus is on developments that not only improve workflow but also improve image quality to ultimately improve the patient experience and outcomes.
To remain competitive, leading players are incorporating AI into their product offerings, pushing solution-led marketing plans. They are also committed to risk-sharing with healthcare providers to engage in partnerships and improve return of investment. Both parties have a vested interest in improving diagnosis and are committed to providing imaging solutions that improve patient outcomes. By focusing on technology that improves patient outcomes, manufacturers are taking on more responsibility in the care continuum. The initial pushback from radiologists is likely to fade as we become more educated on the true potential of AI. IHS Markit predicts that adoption of machine learning and artificial intelligence in healthcare will increase dramatically in the radiology market over the next five years.