The Journey from an Algorithm to a Clinical Solution
Where we are and where we’re going Radiology, as a medical specialty is extremely data dependent. Consequently, radiology is the first specialty to suffer from data overload as well as the first to benefit from the proper utilization of data via novel technologies. In this context there are great expectations for the new generation of radiological AI systems that are likely to impact the day to day radiological workflow.
From its inception, the AI ecosystem has had two driving forces. On one side, were the visionaries that talked about long-term trends. They provided a vision, fostering a nurturing atmosphere and encouraged investment. On the other side, lay the technical experts who focused on the “nittygritty” of algorithms. In retrospect, no progress was possible without collaboration between these two communities.
The time has come to close the gap between these two groups. For the transformation to Next-generation AI to succeed, the AI ecosystem must adopt a holistic view of the entire radiological work process. The AI ecosystem should move from talking about algorithms and models to encompass both the clinical outcomes as well as economic benefits achieved from AI augmented workflow.
To succeed, we believe that the Medical Imaging AI ecosystem will mature into a three-tier system including an Algorithmic, Product and Solution layer. Each layer would address different aspects of the overall solution (and feature different Key Performance Indicators (KPIs). However, only the combination of all the three layers will bring true value to this field.
This whitepaper reviews the current status of AI and explores what it will require to bring the concept of a complete 3 tier solution to fruition. It is designed to help radiologists; informatics experts and other healthcare professionals understand the new direction of AI and how Next Generation AI will benefit radiologists and patients alike.