Karim Karti is the former president of GE Health Imaging and the current CEO of RapidAI – a company founded more than 10 years ago by Dr. Greg Albers, one of the world's leading stroke researchers and director of the Stanford Stroke Center.

For more than 20 years, most in healthcare believed doctors had less than three hours after a stroke to provide treatment. However, Dr. Albers' landmark research ultimately demonstrated that a thrombectomy (a procedure to remove blood clots) as late as 24 hours after stroke onset still benefited patients.

Albers and Dr. Roland Bammer founded RapidAI to streamline the stroke workflow and get patients to treatment faster.

Since then, their AI technology has been applied beyond stroke treatment, to aneurysm, pulmonary embolism and more. Today, RapidAI is being deployed in more than 2,000 hospitals in more than 100 countries, the company reported.

With his extensive experience in medical technology, Karti joined RapidAI this year as CEO with the goal of expanding the company's AI product offerings and growing the company globally.

Healthcare IT News spoke with Karti to discuss AI and the treatment of strokes, the pairing of AI and emergency care, and the potential of AI to save lives and improve outcomes for the millions of people who suffer from common vascular and neurovascular diseases globally each year.

Q. Please explain how the founder of your company was able to use artificial intelligence technology to change the treatment of strokes.

A. For decades, clinicians generally believed they had less than three hours after the onset of a stroke to provide treatment. However, Dr. Greg Albers' research ultimately demonstrated that a thrombectomy as late as 24 hours after stroke onset still benefited patients. These findings completely revolutionized the way neurologists thought about and treated the condition.

But Dr. Albers did not want to stop there. He and his colleague Dr. Roland Bammer knew that time was still of the essence when it comes to stroke, and they wanted to help physicians more quickly diagnose stroke and plan the course of treatment. Together at Stanford University they developed a fully automated, AI-powered image processing software for CT and MRI scans.

Since then, countless studies have shown the efficacy of the AI to support more accurate diagnosis, decrease time to treatment, and ultimately improve patient outcomes. Not only that, but there are hundreds, if not thousands, of clinicians using it today that have spoken to its value. There are even more stroke patients who have benefited from the AI and are alive and healthy today because of it.

Q. How is AI and emergency care a good pairing?

A. AI is and will be helpful in many areas across the healthcare continuum. The urgency and care coordination needs in the ER make AI and mobile workflow technology incredibly helpful and impactful when it comes to decision-making for physicians and care teams.

I'm certainly not talking about replacing physicians with AI – but rather, how AI can better support and reaffirm clinicians during the clinical decision-making process. This is extremely important in emergency care, partially because the ER is multidisciplinary with many different care pathways.

AI can help clinicians triage patients more quickly, choose which treatment pathway makes the most sense for that patient, determine if they have the right capabilities or need to transfer the patient, and ultimately get patients to treatment faster.

AI is also particularly valuable in emergency care because of the speed at which some of these decisions need to be made.

For example, when a patient enters the emergency department presenting with symptoms of stroke, having the right technology to rapidly process scans, produce easy-to-interpret images, support quick diagnosis, and notify and assemble team members in a timely manner can be a matter of life or death – allowing team members to share images and scans easily, helping team members communicate securely and efficiently through mobile technology, and sharing additional information about the patient's medical history – all of which are critical in stroke care.

While we started in the treatment of stroke, we continue to expand into other areas of emergency care that can greatly benefit, such as pulmonary embolism and other vascular and cardiac conditions.

Q. What is the potential of AI to save lives and improve outcomes for the millions of people who suffer from common vascular and neurovascular diseases globally each year?

A. When a patient has a stroke, their brain is not receiving the oxygen it needs, causing the brain tissue to die. So, the faster the patient can get to treatment, the more of their brain will be saved. AI is completely changing the speed at which treatment decisions are made, therefore reducing the amount of brain tissue lost and improving the patient's ability to walk out of the recovery room without life-altering injuries.

Another common condition is brain aneurysm or bulge in a blood vessel that, if it grows large enough, poses a risk of rupturing. When an aneurysm ruptures, it causes bleeding in the brain, also called a hemorrhagic stroke, which can be life-threatening. Identifying aneurysms that are at risk of rupture is key to preventing these severe complications.

When an unruptured aneurysm is detected on a CT scan, neurologists will measure it. However, due to the uneven shape of aneurysms, it can be difficult to get an accurate reading. AI has been proven to provide more accurate measurements and get patients life-saving treatment before it is too late.

Pulmonary embolisms are blockages in a pulmonary artery of the lungs. They can be incredibly difficult to assess and diagnose – not only because of the variety of symptoms that can occur, but also because there is no one single test to detect them.

Treatment plans are also complex and vary based on the patient. Given the amount of time and resources needed to diagnose, assess and treat PE, AI can be key in streamlining this process and improving patient outcomes.

It is important to keep in mind, though, that not all AI is created equally. Amidst the excitement around the potential of this type of technology, unfortunately the importance of clinical evidence has been often overlooked. Strong clinical evidence not only creates trust in the technology for both patients and providers, but is critical to increasing adoption.

AI supported by peer-reviewed clinical evidence is truly the key to propelling the field into the next generation of healthcare.

While we have already seen the impact of clinically validated AI in reducing patient morbidity and mortality rates in these three disease states, we're only at the beginning, and the future is incredibly promising. The opportunity to use AI across all vascular conditions, and beyond, is real and it is here.

The innovations happening across the globe will not only have an immense impact on patient lives, but also on hospital staff, families and caregivers. As we continue to innovate and improve our AI, physicians and hospital staff will have even better support to reduce burden and improve patient outcomes. Everyone wins in that scenario.

Q. You've reported that your company's technology is being deployed in more than 2,000 hospitals in more than 100 countries. What are a couple of actual examples of how the technology improved patient outcomes?

A. We are incredibly proud that there are thousands of patients who have benefited from our technology, but there are, of course, some particularly compelling examples that come to mind. One is a young patient, a 29-year-old firefighter and paramedic with no known pre-existing conditions, who suffered a stroke while working out.

The neuro-interventional team was alerted, so when the patient arrived at the hospital, a CT scan was performed, and an automated notification was sent to the team as soon as the scan results were available. The scan revealed the need for an immediate mechanical thrombectomy (a clot-retrieval procedure that would remove the blood clot causing the stroke).

Because of our technology, the entire team was ready to go right away, and the patient received immediate treatment, restoring blood flow in under 60 minutes.

A second example is a 70-year-old patient who was suffering symptoms of a stroke, including inability to speak. When he was brought to the hospital, they conducted non-contrast CT and CT angiography scans, but neither scan showed that anything was wrong. Doctors then conducted a CT perfusion scan, and, using the AI platform, were able to detect areas of the brain that had been affected by reduced blood flow – pointing to a stroke.

The patient was suffering a distal vessel occlusion, which is a blockage in an artery that is so small that it was not seen on other types of brain scans. The AI helped the doctors determine the appropriate treatment, and ultimately the treatment was successful. The patient regained his speech the next day and had virtually no remaining complications.

Our software has even been used to help treat medical emergencies outside the realm of traditional neurovascular and vascular conditions. There was a recent story of a boy who had been impaled by a metal straw, which struck an artery in his brain.

One of our modules, which is typically used to look for salvageable brain tissue, was used to ultimately determine that the patient's carotid (which compromised the entire right hemisphere) was severely hypoperfused. The doctors used this information to guide treatment decisions and save the boy's life.

Overall, we measure our success by the success of providers, payers and physicians, and ultimately impact on patients, which these stats represent:

The technology has been proven to reduce door-to-CT times by 69%, door-to-CTA interpretation by 63%, door-to-decision by 52 minutes, door-in-door out by 60%, door-to-groin by 75% and door-to-needle by 49%.

AI is driving a big revolution. It began with the original research from Dr. Albers, which led to the extension of the treatment window for stroke patients to 24 hours, thereby helping save an enormous pool of patients that would not have been treated otherwise. 

Our ambition is to continue to pioneer the field and advance it into the next generation of healthcare – in which AI-powered decision-making and collaboration are available at your fingertips, anytime, anywhere.

Twitter: @SiwickiHealthIT
Email the writer: bsiwicki@himss.org
Healthcare IT News is a HIMSS Media publication.