Reclaiming time in radiology: how AI can help tackle staffing and care gaps by streamlining workflows
Nov 26, 2024 | 10 minute read
About the author About the author
Shez Partovi obtained his medical degree from the McGill University, in Montreal, Canada and completed his neuroradiology subspecialty at Barrow Neurological Institute in Phoenix, AZ. He is a serial entrepreneur and has launched several health IT companies, including two on telehealth.
You are about to visit a Philips global content page
Continue If only we had more time. It’s a sentiment I felt every day as a radiologist, juggling a relentless stream of imaging studies. Time pressure in radiology has only intensified – with radiologists and imaging staff experiencing mounting strain and patients facing potentially life-altering delays in care. It’s a systemic problem, and in speaking with radiology leaders around the world, they’ve made it clear: the need isn’t just for better imaging quality but also for tools that let them work more effectively with the people they already have. And that’s exactly where AI has the power to help.
“We don’t just need more pixels; we need more productivity.” When I talk to my innovation teams, I often share this insight from a radiology leader I met at a conference to highlight a pressing need we’re hearing from customers worldwide. Yes, radiology leaders value the latest advances in imaging resolution, but above all, they are looking for solutions that can help them boost capacity and gain efficiencies, without compromising on quality. In an age of growing patient demand and persistent staff shortages, time has become the most precious resource in radiology – and there’s never enough of it. The human costs of this time crunch are well-documented. More than one-third of radiologists are experiencing burnout, according to a 2024 study in the American Journal of Roentgenology [1]. The lack of time and staff isn’t just a challenge for radiologists – it’s affecting patients too. In our 2024 Future Health Index report, 77% of healthcare leaders reported care delays due to staff shortages. Backlogs in imaging studies risk missed diagnoses, while delays in cancer treatment – which can be life-threatening – have unfortunately become ‘routine’ in some healthcare systems [2, 3]. We can’t recruit ourselves out of this crisis, and building new radiology rooms has often become prohibitively expensive. Instead, we must empower existing radiologists and imaging staff by enhancing their skills and efficiency. That’s where AI comes in. However, there’s a catch: in a radiology workflow, every step is interconnected – from gathering and organizing prior patient data to image acquisition, interpretation, reporting, and communicating imaging studies to patients. If we optimize just one step in isolation – for example, by speeding up the scanning process – we may only create a bottleneck in another – since, in this example, it means the radiologist now has to read more studies at a faster pace. A piecemeal approach to AI won’t work. We must optimize the entire workflow. Here’s how:
How AI can help radiologists save time and reduce care delays
To apply AI in a meaningful way across the radiology workflow, we must begin with the needs of radiologists, imaging staff, and patients. Imaging interpretation is what radiologists are trained to do and what they tend to find most fulfilling (I certainly did). Naturally, discussing cases with colleagues and interacting with patients is incredibly rewarding, too. From my experience, these are typically the aspects of radiology work that imaging staff enjoy most as well. By automating non-value-added tasks with AI, we can free up time for these core activities – which benefits the patient too, because they receive more personal, timely care. Take CT, one of the most widely used imaging modalities, as an example. During image acquisition, AI can assist with patient positioning, reducing positioning time by up to 23% [4]. An AI-enabled camera mounted above the patient table identifies key anatomical points and orientation, automating what is typically a manual process. Especially for less experienced technicians, this can boost their confidence and save time on positioning, allowing them to focus more on the patient. AI-enabled image reconstruction techniques can then provide high-quality images at high speed while minimizing patient exposure to radiation [5]. And the story with AI doesn’t end here. Smart workflow prioritization can automatically assign cases to the right subspecialty radiologist at the right time in the right order, based on various parameters including AI-assisted screening. As a result, one healthcare organization in Italy was able to speed up its radiology workflows by 50% – significantly increasing its capacity to provide better care to more people [6]. What’s more, AI can support radiologists in their clinical decision-making. For example, in CT lung cancer screening, where early detection is critical for better patient outcomes, AI can help radiologists identify lung nodules 26% faster and detect 29% of previously missed nodules [7]. For patients, this may mean receiving earlier and potentially life-saving care.
The potential of generative AI: from faster reports to patient empowerment
All of what I just described is already possible today. But with the rapid rise of generative AI, we are just beginning to scratch the surface of what can be accomplished. It’s no wonder that, according to our 2024 Future Health Index report, 85% of healthcare leaders are already investing or plan to invest in generative AI within the next three years. Generative AI promises to create further time savings that benefit both radiologists and patients. For example, conversational reporting using generative AI allows radiologists to dictate as they wish while the resulting report is transformed into a consistent reporting format. The AI refines reports in real-time, adds diagnostic impressions, and flags inconsistencies back to the radiologist. This reduces editing time and maintains high reporting quality by integrating patient histories and clinical context. For patients, faster reporting leads to quicker diagnoses and better overall care. In areas like cancer care, generative AI could be a game-changer by summarizing a large number of historical reports to give care teams immediate insight into a patient’s history. It is not uncommon for oncologists to have to review numerous prior imaging reports. Generative AI could provide quick summaries to help them focus on creating treatment plans rather than sifting through extensive reports. I envision a future where generative AI empowers patients as well, by helping to translate imaging studies and other medical reports into layperson’s terms, at scale, and in the language of a patient’s choosing – something that seemed impossible just a few years ago. This could help bridge the health literacy gap that prevents patients from fully understanding their diagnosis or actively engaging in their own care. Imagine being able to talk to your own medical record – with generative AI, this could become a reality in the not-too-distant future.
Advancing AI in an open ecosystem
To fully harness these possibilities of generative AI, we need the scalable computing power of the cloud. Integrating diagnostic capabilities in the cloud will further enhance clinical collaboration between different specialists – including oncologists, radiologists, and pathologists – by facilitating anytime, anywhere access to essential information. Ultimately, this will create one integrated diagnostic environment where all relevant patient data is easily accessible for faster decision-making and more personalized treatment plans, with AI acting as a smart assistant. When you consider all the moments in the radiology workflow that can be optimized with AI, and all the strands of patient data we need to bring together, it’s clear that no company can achieve this alone. Partnerships are essential – whether it’s by enabling different systems to exchange information more seamlessly, or by integrating AI algorithms into common technological platforms. Collaboration with regulatory bodies is equally important to accelerate the safe and responsible adoption of AI. As healthcare leaders, we have the opportunity to embrace AI in a way that empowers both healthcare professionals and patients – creating a future where radiology is defined not by constraints, but by the quality of care we can help provide. With radiology departments under increasing pressure, we certainly have no time to wait. To learn more about how Philips is partnering with healthcare leaders to help them advance precision imaging for more patients, join us at the upcoming RSNA annual conference. You can also follow @PhilipsLiveFrom for updates from #RSNA24.
