Healthcare’s Challenge – and Opportunity – to Scale AI in Imaging

October 5, 2020

Invest wisely in a strong foundation to ensure continued innovation.

The pandemic is forcing healthcare organizations to make hard decisions when it comes to managing operations and making investments across the enterprise. This includes the realm of radiology, where clinicians are stretched thin in the best of times and even more challenged in the time of a pandemic. Healthcare leaders know the answer is to innovate, but also recognize the need to balance financial realities and budget constraints. They have a challenge – and opportunity – to invest wisely. It’s never been more critical to focus on AI innovations that can yield both immediate short-term and longer-term value.

Anyone who has entered a healthcare facility in recent weeks can immediately see that things are very different – from social distancing, patients and providers in masks, and sparsely populated waiting rooms. Waiting rooms are not just sparsely populated due to COVID social distance restrictions as individuals continue to defer care. New research from TransUnion Healthcare, a subsidiary of TransUnion focused on healthcare revenue protection solutions, finds that use of services remains low, with emergency department visits down 25 percent the week of Aug. 16 compared to pre-COVID volumes – making it the hardest hit department. Inpatient volumes were down 9 percent. Many facilities are experiencing extreme financial pressure; operating margins at hospitals are down 96 percent through July, according to healthcare consulting firm Kaufman Hall. This forces immense workload pressure on many clinicians, including radiologists, who remain on skeleton crews.

The pressure on staff will continue to mount further as facilities eventually begin to work through an unprecedented backlog. For example, an estimated 40 million mammography procedures are performed in the United States annually. At the height of the pandemic, the need to delay or cancel appointments equates to nearly 770,000 delayed mammography procedures per week. As patients start to resume normal care, radiology staff must reschedule procedures efficiently and safely, optimizing their resources as best as possible.

The current situation adds complexity to the specialty when we are already heading toward a shortage of radiologists in the United States. The overall demand for radiology services has been steadily rising since 2013 as the population ages. Nearly half of the radiologists in the United States are of retirement age, and fewer radiology residents are waiting to fill these soon-to-be-vacant positions. By 2033, the United States could have a shortfall of nearly 42,000 radiologists and other clinical specialists, according to the Association of American Medical Colleges.

In short, radiology resources are overwhelmed at a time when specialized staff is needed most – increasing the likelihood of clinician stress and burnout, as well as complications in patient care. Healthcare organizations need help to weather the current storm and, hopefully, thrive in the new normal. They must take the opportunity to invest wisely today to address these immediate challenges, as well as set the stage for use of innovative technologies like imaging AI at scale – to drive improved care and greater productivity. This is where teleradiology and expanded use of artificial intelligence (AI) show promise. But, all of this requires a strong infrastructure and foundation for the data at the heart of diagnostic imaging.

Teleradiology in the Age of COVID and Beyond

Teleradiology, unlike telemedicine before the pandemic, was already deeply rooted. The market was $1.2 billion in 2019 and is likely to continue to expand for reasons related to and beyond COVID. In addition to the talent shortage and greater demand for specialized modalities, teleradiology is set to expand to $2 billion by 2024 due to the longer read times required for specialized modalities, demand for always-on diagnosis, and greater adoption of cloud-based technologies.

A January 2020 survey shows half of radiologists are burned out, with “too many work hours” among the top causes. Organizations need to be able to power greater productivity to deal with immediate requirements and set the stage for the future. Teleradiology provides the opportunity to better support radiology staff with speed, accuracy, and decision support. Wise organizations recognize the opportunity to preserve care and also optimize the staff they have. Of course, teleradiology alone does not ease the incredible burden and pressure on radiologists. Organizations need to work smarter with technology solutions enabled by machine learning and next-generation AI.

The AI Connection

COVID-19 has accelerated the need for enterprise imaging and telehealth via remote access to imaging and reports. It also shines a spotlight on the need for expanded use of AI and machine learning in imaging, with the dual goals of improved diagnoses and greater clinician wellness and job satisfaction. It can also have a positive impact on an institution’s financial health. Industry studies estimate that 37 percent of a hospital’s revenue is derived from imaging. More than ever, facilities need to optimize the use of their imaging facilities to protect the bottom line.

Without question, AI can and must play a growing role regardless of whether radiology is conducted onsite or via teleradiology. We, first, think of AI in the clinical setting, where it serves as a support for diagnosis – more as augmented intelligence to improve accuracy rather than a workforce replacement. AI also has the potential to reduce bias and inform clinical decision-making and treatment, while decreasing turnaround times. Applying AI on less complex modalities can also free up radiologists’ time to focus on specialized modalities, including greater detection and visibility of soft tissues, cancers, and tumors, and reduce burnout.

Further, AI can improve explainability and transparency in imaging, and ultimately support integrated care initiatives, which rely on swaths of data analytics at the point of care to be actionable. For example, in a major milestone, Centers for Medicare & Medicaid Services recently approved Viz.ai, an AI stroke platform, for reimbursement, enabling hospitals to widely adopt the advanced technology for improved stroke care – and representing the first AI software to secure this approval. Imaging tech that delivers on AI, teleradiology, and security while ensuring room for future innovations will be essential.

Invest to Optimize Today and Tomorrow

To accelerate AI, healthcare organizations need to create a modern data experience. They already got a taste of what it’s like to innovate faster, as pressured by the pandemic and quick shift to telehealth. But now, six months later, healthcare organizations are grappling with immense financial pressures. They want to continue to accelerate investment, but they must invest smarter than ever during austere financial times. Enabling a modern data experience is an investment that yields both short and long-term dividends.

The foundation they create for AI matters. To ultimately support AI at scale, organizations seek stronger and better strategies to store and access their data quickly, easily, and affordably. This all starts with the data foundation and data availability: Healthcare organizations need to ensure a modern data experience powered by a data-centric architecture that works in real time, prioritizes automation, is reliable and secure, supports the multi-cloud, and allows for constant innovation with a goal of lower TCO.

Organizations need a strong investment that delivers today and well into the future. The modern data experience not only sets the stage for expanding AI use, but provides the foundation to support new technologies, larger volumes, and speed that organizations need today. It provides not only the opportunity to solve immediate challenges, but also support continued, accelerated innovation.

What Is the Modern Data Experience?

A Modern Data Experience enables healthcare organizations to extract maximum value from their imaging platforms while reducing complexity for radiologists and expense – all in service of value-based care. It starts at the data foundation, with a data-centric architecture, and has four main characteristics: simple, sustainable, fast, and seamless.

The modern data experience is simple – providing an easy clinician and patient experience by delivering the image to the right person at the right time. It is also future proof – eliminating forklift upgrades and the cost of replacing legacy technology to bring imaging into the not too distant future of teleradiology and expanded use of AI. It’s not easy to predict storage needs, but the right data foundation enables strained healthcare organizations to use only what they need and scale up or down accordingly, even as the organization’s business needs evolve.

With the explosion of imaging data – to 2 trillion images per year globally – radiology’s data foundation must be on the leading edge of performance. A modern data experience enables increased throughput and drastically decreased turnaround time that overworked radiologists desperately need. It is reliable, consistently delivering needed performance no matter the workload conditions or a specialist’s location. Moreover, it is seamless, consolidating applications and removing data silos that have become so common.

Given these pressing workforce trends and the potential of AI and teleradiology, facilities know the answer is to innovate while navigating difficult financial realities. Therefore, the key is to invest wisely to lay a strong data foundation – one that will aid clinical operations immediately but serve as the crucial long-term investment to fuel future innovation and weather any future challenges.