By Luke Bideaux, BSRT, CIIP

Click to Read on AuntMinnie.com

August 12, 2019 — A renewed level of excitement is sweeping the field of health imaging due to advancements in imaging-based artificial intelligence (AI). In 2018 alone, over $500 million in capital investments was provided to companies developing AI-enabled medical image analysis solutions, according to Signify Research. There’s no doubt that AI has the potential to dramatically disrupt the field of health imaging. As more mature implementation models begin to take shape, organizations are developing strategies to leverage AI as a supplemental tool for their clinical staff to bring about increased efficiency and improved quality.

Resources with specialized skills in imaging informatics are key to the development of AI strategies that will have the greatest impact on improved organizational efficiency and quality of care and that are aligned with overall enterprise imaging strategies. As I’ve done with part 1, part 2, part 3, and part 4 of this series, I will demonstrate in this final article the value of certified imaging informatics professionals (CIIPs), specifically relating to how they can help spearhead these AI initiatives and achieve a multitude of positive outcomes.

The cornerstone of your AI strategy

It’s important to understand that AI encompasses any machine behavior involving the completion of tasks based on a set of specified rules to solve problems. Based on that definition, AI is nothing new in medical imaging. For instance, workflow engines that present particular studies to appropriate personnel based on rules are considered AI. Hanging protocols that decipher the images within a study and present them in a predefined manner are another example. While AI is not new to medical imaging, significant developments are occurring, such as algorithms that can help providers diagnose certain pathology with significantly reduced variability and reading times. These exciting breakthroughs are encouraging healthcare organizations to develop AI strategies that can help them take advantage of these advancements as they head into the next generation of medical imaging.

Developing an AI strategy should be a component of an institution’s larger enterprise imaging strategy. These types of strategic initiatives are typically governed by steering committees or governance teams. For instance, an organization could form an AI Governance Committee, which could be set up as a subcommittee of the Enterprise Imaging Governance Committee. Regardless of the structure, it is critical that the appropriate personnel be involved in the development of an AI strategy.

Certainly, clinical leadership — including radiologists, cardiologists, and/or other clinical personnel who will use the AI technology — will play an important role in helping to define an organization’s exact needs and determine the clinical significance of any proposed solution. IT and operations leaders must also be involved to collaborate on the development of the technical architecture and workflow models to support the AI algorithms. Most importantly, a CIIP should be identified to serve as the cornerstone of the AI committee.

CIIPs are board-certified through the American Board of Imaging Informatics (ABII) in knowledge bases such as medical imaging workflow optimization, medical imaging standards, electronic medical record (EMR)/RIS/PACS integration, strategic planning and reporting, and many more applicable areas. As such, CIIP resources would allow the given committee or team to move from high-level discussions about AI to development of a strategic plan, including such critical details as technical framework development and integration design.

To support AI initiatives, CIIP resources may be found internally or may need to be brought in from outside consulting agencies. For organizations looking to develop their own internal CIIP resources, the Society of Imaging Informatics in Medicine (SIIM) provides opportunities for professionals of various backgrounds (PACS administrators, radiologists, IT staff, etc.) to educate themselves in imaging informatics and prepare for certification.

In previous articles in this series, I have demonstrated examples that many readers can relate to regarding the successes that are possible when the right CIIP resources are in place, as well as the failures that can result from lack thereof. In this fifth and final article of the series, we will shift gears from studying past lessons learned in the field of imaging informatics to predicting the effects that CIIP resources will have on imaging practices’ use of AI in the future. Join me as we gaze into our crystal ball to see how AI technology will affect Adventure Health — our fictitious imaging practice from parts 1 through 4 of this series — depending on whether it chooses to utilize proper CIIP resources or not.

Option 1: CIIP

The year is 2024. Five years ago, Adventure Health began developing an AI strategy that would leverage both the outstanding expertise of its clinical staff and advanced AI technology to increase efficiency and improve quality of care. The organization identified CIIP resources to collaborate with its clinical experts, IT, security, and operations leadership staff to evaluate AI applications that would result in meaningful improvements in efficiency and the delivery of quality care. Together, the team made recommendations based on performance, integration capabilities, system architecture, and potential impact on clinical workflows. Fast forward five years, and Adventure Health is now enjoying the benefits of its hard work after procuring and implementing appropriate AI solutions.

In radiology alone, read times were reduced by an overall average of 42%. In specific areas, such as CT lung cancer screening, read times were reduced by as much as 80%. These efficiency gains would not have been realized without the valiant efforts of the CIIP resources who guided the AI strategic planning efforts. They were instrumental in the development of workflows that incorporated AI technology in ways that saved Adventure Health radiologists clicks, rather than adding them.

These workflows allow studies to be automatically processed by the appropriate AI algorithm based on structured data elements of the study ordered from the electronic health record. Studies are designed to appear on radiologist worklists for review only after the appropriate AI processing has been performed. With a single click, radiologists can incorporate the findings from the AI software into their final reports. Radiologists also have the opportunity to modify these findings in a manner that will train the AI application to learn from their input. Looking back, Adventure Health’s radiologists are amazed when they think about how far things have come.

The reading room isn’t the only area where AI has made a significant impact. Technologists involved in generating 3D visualizations, bone removal procedures, maximum intensity projections (MIPs) and other types of advanced visualization tasks have also witnessed their workflows change dramatically over the past five years. Processes that used to take up to 30 minutes to complete can now be accomplished within seconds. These efficiency gains have enabled Adventure Health to repurpose its technologists and reduce its reliance on per diem technologists by as much as 75%. With a solid foundation for further AI expansion, along with the appropriate resources to support these developments in place, Adventure Health is excited about what improvements the next five years will bring.

Option 2: No CIIP

The year is 2024. The past five years have been a challenge for Adventure Health due to its inability to keep up with the ever-changing technologies in medical imaging. Without the resources to develop an appropriate AI strategy, Adventure Health failed to develop an appropriate roadmap for implementing AI. Instead, Adventure Health has haphazardly implemented an assortment of AI algorithms without regard for tight integration with its existing systems.

The radiologists at Adventure Health have seen AI significantly impact their workflow but, unfortunately, not in positive ways. Radiologists are forced to log in separately to various web portals to review the findings that their AI software provides. Without any meaningful integration, radiologists must manually dictate any findings from the AI solution into their final reports. Radiologists complain that the lack of integration has made use of AI technology cumbersome. Resultantly, many Adventure Health radiologists avoid using the technology altogether. Technologists at Adventure Health are responsible for manually pushing studies to the appropriate AI application, which has actually added steps to their workflow.

Due to a lack of CIIP resources to provide strategic guidance, integration with Adventure Health’s AI solutions is virtually nonexistent. Unfortunately, this means that the techs are required to utilize a complex process that involves exporting images from PACS and importing them into various AI-related web portals. To keep up with the workload these extra steps have created, Adventure Health was forced to hire two additional techs. Looking ahead, Adventure Health’s new five-year plan involves making several cutbacks, and leadership wonders why AI has not lived up to expectations.

In the year 2024, failure to utilize CIIP resources resulted in the following costs for Adventure Health:

• Opportunity cost of 42% radiologist efficiency gain: $318,000
• Opportunity cost of 75% reduced reliance on per diem technologists: $82,000
• Two technologist full-time equivalents (FTEs): $118,520

Total: $518,520

Running total of one-year costs, from parts 1 through 5 of this series: $1,489,096

Additional negative outcomes from failure to utilize proper CIIP resources include the following:

• Increased radiologist frustration, FTE requirements, and system logins
• Reduced workflow efficiency, use of AI technology, and quality of patient care

It’s clear that organizations will need to dedicate resources to support new AI initiatives over the next five years. Organizations that choose to engage CIIP resources throughout the process will have the best chance of achieving success with AI in the future. Now is the time to lay the foundation for AI to make a positive impact on your organization for years to come. Be sure that your team is equipped with the resources needed to embark on this important journey.

This concludes my series on the value of CIIPs. I’ve demonstrated their value in numerous ways, resulting in significant cost savings, efficiency gains, end-user satisfaction, and more. As I wrap up this series, I would like to thank all CIIPs for investing in the professional development required to attain and maintain their certifications. Your dedication to continued development as imaging informatics professionals is a testament to your commitment to improving patient care within your community.

My message to healthcare organizations: Do not miss out on the value of CIIPs! As you perform your resource planning, don’t forget to consider the costs of not utilizing CIIPs, along with the benefits of using them. While financial considerations are always important, remember to consider the impact CIIPs have on other areas of your operation such as radiologist morale and patient care.

My message to CIIPs: Demonstrate your value! Don’t rely on others to do it for you. Show leadership the significant impact your efforts have made on your organization. Take ownership of your career by continuing to improve your value through education and professional development opportunities.

Luke Bideaux is the founder of Vega Imaging Informatics, a consulting agency offering managed services and software solutions related to imaging informatics.  If your organization is in need of AI consulting services, or any imaging informatics-related service/solution, please contact us at [email protected].