Courtesy: GE HealthCare
Ultrasound imaging is one of the most used diagnostic procedures in the healthcare field. The technology plays a critical role in timely diagnosis, disease management and treatment planning across many medical specialities, from paediatrics to women’s health.
Advances in ultrasound technology, such as embedded AI and deep learning algorithms, can significantly assist providers and the patients who rely on them for the best possible care. Clinicians can make confident diagnoses faster as equipment becomes more innovative, meaning the patients they treat get the right interventions at the right time.
Strategically positioning ultrasound as a valuable diagnostic tool in a provider’s digital ecosystem of connected medical devices can take ultrasound to the next level with access to analytics and productivity solutions.
“As we think about the pressing problems of increased demands, the need for more efficient productivity and the goal to improve health outcomes exists not only in ultrasound, but across the global healthcare spectrum,” explains Karly Yoder, General Manager and Chief Digital Officer for GE Healthcare Ultrasound. “At GE Healthcare, we think it requires a strategy and a focus on precision health. It means we want to do the right thing at the right time for every patient at a global scale. That is a big vision, but we know it’s the right vision if we want to move healthcare into the future that it requires to address these pressing issues.”
Ultrasound is used across nearly every clinical area in healthcare. Industry leaders such as GE Healthcare are innovating ultrasound with improved usability and flexibility, as well as improvements in ultrasound’s image quality and clinical decision support tools. Thanks to important artificial intelligence (AI) tools, clinical applications such as breast ultrasound and pediatric interventional radiology are seeing some impressive results.
AI tools, connectivity, and innovation in ultrasound
Interest in AI in healthcare is on the rise, with $4 billion invested into the sector in 2019, up from $2.7 billion in 2018. Many AI-based solutions have already been adopted into healthcare environments to support specific processes and tasks. There are also many other new AI solutions and machine learning algorithms available, but the adoption rates don’t reflect the rates of development.
“The problem is when we actually look at adoption of AI for some of these digital technologies, we see a rate of about 10 to 15 percent of adoption,” explains Yoder. “Why is that? We as a healthcare industry need to be obsessed and focused on how we make this adoption simple, basically invisible, into the workflows that healthcare providers already use and trust across their departments, and across their institutions.”
Clinicians are less likely to be enthusiastic about AI solutions if they cause them more administrative work in their workflow. To gain physicians’ trust, according to Harvard Business Review, AI software developers will have to clearly demonstrate that when the solutions are integrated into the clinical decision-making process, they help the clinical team do a better job. The tools must also be simple and easy to use.
“GE Healthcare is approaching AI development and implementation from two perspectives,” says Yoder. “The first ‘mile’ is all about building applications that matter to customers and driving operational as well as clinical outcomes in the right direction. Getting this right means leveraging AI and the building blocks of a healthcare ecosystem. The last ‘mile’ is about addressing the adoption challenge. So, if you’ve built a wonderful solution that does incredible things in the healthcare setting,” she explains, “but you’ve developed it in a way that requires a clinician to go to a new computer, open a new application, or get on the phone and have to call somebody, you’re not going to see adoption because it breaks the workflow.”
How AI is changing the game in breast ultrasound
To aid clinicians in finding breast cancers, AI-based tools have been introduced and show great promise in increasing diagnostic accuracy, as well as creating substantial productivity improvements.
An AI-based quantitative risk assessment tool, Breast Assistant, powered by Koios DS, provides clinical decision support for physicians and technologists when using ultrasound to detect and diagnose breast cancer. It’s easy to use, and results have shown higher rates of diagnosis and fewer biopsies.
“We’ve got issues of missed cancer. We’ve got issues of avoidable biopsies. We’ve got issues of unnecessary follow-up and a dramatic amount of variability,” explains R. Chad McClennan, CEO of Koios. “Now there’s a new solution. There’s diagnostic AI.”
He explains that their AI solution is not like traditional CAD. It’s not the binary choice between normal and abnormal. It’s the interpretation of something abnormal. Providing clinicians with an automatic BI-RADS assessment and accurate interpretation results in a positive impact on patient outcomes, so that there’s no need for the patient to stress or undergo a subsequent procedure in some cases.
Improving outcomes using point-of-care ultrasound in pediatric IR
In other clinical areas such as interventional radiology (IR) and urgent care, ultrasound is well established as a safe and effective imaging modality for the rapid diagnosis and management of many emergency conditions. At the bedside, it also improves success and patient safety during invasive procedures. With many technological and digital advances as well as the increasing availability of imaging technologies, there has been considerable expansion of the use of clinical ultrasound—including both radiology-performed consultative studies and point-of-care ultrasound (POCUS) studies.
Jay Shah, MD, pediatric interventional radiologist (IR) at Children’s Healthcare of Atlanta and Assistant Professor in the Emory University School of Medicine Department of Radiology and Department of Paediatrics in Atlanta, Georgia, covers two hospitals with a group of four IRs. They are tertiary care hospitals with a collective 800 beds. Two hundred of those are ICU beds, including NICU, PICU, TICU and any other acute care.
Ultrasound is a mainstay for Dr. Shah and his team, and in addition to its clinical value, it helps them reassure their patients’ parents, most of whom are a little nervous or a little anxious because they’ve usually seen several other providers before landing in pediatric IR. Due to the types of clinical issues they treat, they often use ultrasound for biopsies, aspiration and drainage, but because it involves no ionising radiation, it can be utilised more.
“Ultrasound is our workhorse,” notes Dr. Shah. “It’s real-time. It’s accurate, and it gives off no ionizing radiation. Additionally, it’s widely available. And in [pediatrics] we try to use it as strategically as we can. We successfully utilize it as a therapeutic procedure whenever it is available. In patients with repeated procedures, it offers no radiation exposure when we do follow ups, and we can monitor lesions that otherwise may or may not require retreatment.”
The team relies on ultrasound for procedures such as tumour ablation, where they can use central venous access at the bedside or in emergent situations. They also perform thoracic interventions and vascular malformations, as well as use advanced techniques such as contrast-enhanced ultrasound.
“For us in the clinic, we use it just like every other clinician does. We take a thorough history, and we do a good physical,” explains Dr Shah. “But the other thing we get to add is what I like to call the radiology physical, which is point-of-care ultrasound. It’s so crucial for us, and it helps our parents leave with a definitive plan. We can sometimes give them a preliminary diagnosis, but even if we can’t, we can give them some sort of idea as to what’s going to come next and whether or not the patient’s going to need further imaging or a procedure.”
Driving a smarter, more efficient future with ultrasound
Ultrasound technology and innovation are evolving to meet the needs of the changing healthcare landscape. The need for more efficient solutions is driving ultrasound providers in the industry to create smarter solutions. AI and deep learning technologies provide clinical decision support, save time, and importantly, create solutions that are easy to use and implement across health systems. GE Healthcare continues to support clinicians with game-changing solutions in ultrasound, such as powerful portable ultrasound for point-of-care use, and deeply embedded AI tools for accurate breast cancer detection.
