Medical imaging is one of the fastest growing fields of medical care. The field has come a long way from the grainy black and white images we relied on just a decade a go to diagnose everything from broken bones to the presence of tumors. In the last decade medical imaging has evolved to produce rich, multi-dimensional, high resolution images that help clinicians diagnose illness more quickly and more accurately to deliver better patient outcomes.
With these new imaging techniques providing clearer insight into what is happening within the body, the amount of data that is generated by medical imaging departments is creating an additional opportunity for clinicians to advance medical care. With multiple care areas creating massive amounts of medical information, this vast data repository has created a unique opportunity for clinicians to leverage the power of artificial intelligence (AI).
We spoke with Duleep Wikramanayake, Chief Information Officer for SimonMed Imaging and Tony Turner, Strategic Partner Manager for Healthcare at NetApp to find out more about why medical imaging has become one of the most compelling use cases for AI and how this technology will benefit both clinicians and patients.
AI Uses Case in the Field of Medical Imaging
“As medical imaging modalities improve the precision of the images they’re gathering, we are discovering things that the eye can’t see,” said Wikramanayake. “Your eye can only see so much; with AI, computer-aided diagnostics is looking even deeper from a pixel-to-pixel level. That detail, layered with retrospective analyses and related diagnoses, gives a view we’ve never had before.”
While Wikramanayake says that AI is helping guide healthcare toward better diagnostics, he thinks its use can go beyond that. “I think we need to leverage AI to dig deeper and look to a cause,” he said. “It is almost impossible for a human being to do all that research and collect that data, so this is where I think AI will make the greatest difference.”
In fact, Wikramanayake hopes to soon see AI as something that is a standard component built into medical imaging, rather than used “after the fact,” as it is now. With AI built in, an anomaly could be spotted, and further exploration could take place during the exam. It also could improve how the technician conducts the exam, speed up the time to delivering images for the AI to interpret, and help the radiologist confirm insights in the moment for faster, more accurate diagnoses.
“What if, at the application level, AI could issue an alert and notify the clinician if an anomaly was detected? Or what if a healthcare organization could create a database of patient information for AI that would enable the technology to associate symptoms with a diagnosis. Because AI is a self-learning technology, the system would ‘know’ which symptoms are, or aren’t, related to a specific diagnosis. Not only would this help the clinician diagnose disease more quickly, but it would also have a positive impact on patient outcomes.”
Building the Case for Investment in AI-Ready Infrastructure
According to Wikramanayake healthcare organizations must be willing to invest in the IT infrastructure, including the computing power, that is needed to support AI. Right now AI is expensive, which presents barriers to many healthcare systems. However, he predicts that this will change in six months to a year.
All of this technology – AI, automation and whatever comes next – has to help the organization from a patient care and an ROI perspective. “One of the biggest hurdles in AI for hospitals without the proper infrastructure are multiple silos of clinical data. This makes it hard to gain access to all that valuable data to train AI algorithms,” added Tony Turner, Strategic Partner Manager for Healthcare at NetApp.
“IT infrastructure is incredibly important. Without it, you can’t successfully implement AI. It’s impossible,” Wikramanayake explained. “And when I say IT infrastructure, I mean everything that goes with it: security, storage, data management, and all the tools that govern compliance around the data,” he shared.
Harnessing the power of AI to improve patient outcomes is within reach for most healthcare systems in the United States. While there are several different pathways to demonstrate the vitality of AI to clinical care and patient outcomes, building and implementing use cases in medical imaging represent a ‘quick-win’ opportunity for clinicians and healthcare information technology specialists.
Both the quality and quantity of data needed to fuel AI are already present in the medical imaging field. With petabytes of data readily available for analysis and learning, there’s a real opportunity to deliver results that will benefit both patient and clinician quickly. From augmenting the ability of clinicians to read images, to reducing time to diagnosis, eliminating errors, and identifying additional clinical information to further refine treatment, the use cases for AI in medical imaging are clear.
What’s needed now is for healthcare systems to make the necessary investment in their IT infrastructure to harness the potential they hold in these stored images. “The supportive infrastructure must be in place and capable of scale before you can experience the power of AI,” Wikramanayake concluded.