Making AI meaningful - How AI is Being Used to Transform Radiology Practices

With the increasing demand for radiological imaging, along with the continuous growth of image density and complexity, radiologists now more than ever working under the increased pressure of time

This pressure can no longer be compensated by improving traditional PACS tools and technologies such as workflow automation, speech recognition and image pre-processing. The impact? Lab and radiology test turnaround times, delaying patient disposition, have been identified as the main causes of reduced ED throughput.[1]  And as we have become acutely aware of with the current COVID-19 epidemic, overcrowding stresses the ability of our Emergency Departments and ICUs to provide immediate access and stabilization to those patients who have an emergent medical condition.

All this is providing for a tide of interest in incorporating Artificial Intelligence (AI) into reading workflows, for it has been shown to help:

1. ‘Intelligently’ automate manual tasks, freeing time for the radiologist to focus on supporting informed decision-making and personalized care,
2. Provide decision support to radiologists – notifying of emergent conditions and providing intelligence to back that up. This includes both the read criticality as well as information to support a more narrow and accurate differential diagnosis, and
3. Assist in detecting patterns in image data that are too subtle or complex for humans to easily see.

In other words, AI can assist radiologists with not only managing a growing volume and complexity of exams, but also ensuring they are always reading for the right patient at the right time – to help provide a finer separation between STAT and Critical. And to assist in creating a richer more definitive report that can help enable care teams to provide the best possible outcomes for their patients.

This is why GE Healthcare has created the Open PACS AI experience. It allows our Centricity™ PACS customers to incorporate and utilize AI results coming from a wide variety of sources:

• It can come directly from a scanner or an edge device connected to a scanner
• It can be acquired from GE Healthcare or a wide variety of third-party AI vendors
• It is orchestrated by a flexible and scalable workflow engine that is part of the PACS
• And it is flowed into your radiologist’s existing reading workflows through a uniform and seamless experience in both the Centricity Workflow Manager and Universal Viewer

This platform approach, enabled by a flexible and open orchestration engine, can facilitate seamless integration of AI based clinical applications without requiring organizations to change their workflows.

It allows easy onboarding of existing or new AI applications from multiple vendors – providing rapid and robust scalability, it provides a uniform interface and user experience across multiple AI based applications, and most importantly, it does not add any additional clicks for the radiologist.

The reality that healthcare today is changing at a rapid pace. In radiology, it’s not enough to read faster. Nor is it realistic. Therefore, today’s environment is one where AI is no longer an idea - or the future – it’s a requirement.   

While this evolution may change the radiologists’ role, it also means they can spend less time performing manual tasks – and more time interacting with critical members of the care team and their patients. Therefore, the introduction of AI brings with it an opportunity - An opportunity to transform imaging practices to enable higher quality of life and improved outcomes for radiologists as well as patients.  

[1] Jomon Aliyas Paul, PhD, Li Lin, PhD, "Models for Improving Patient Throughput and Waiting at Hospital Emergency Departments," Administration of Emergency Medicine| Volume 43, ISSUE 6, P1119-1126, December 01, 2012