Article

Innovation and Integration: Making Healthcare Intelligent with AI

To be precise means to be focused on the details—focused on the specific situation to be accurate and effective. But in healthcare, the idea of “precision health” is carried out on a global scale, not in just one instance. Specifically, precision health means to be able to do the right thing at the right time for every patient on a global scale. But how is that achievable? At GE Healthcare, with our Edison® Ecosystem, we believe data analytics and AI are key to success.

Efficiency that doesn’t compromise quality

To take a look at the impact of AI in healthcare, let’s consider prostate cancer—1 in 9 men are diagnosed, and 1 in 41 will die of the disease[1]. Radiologists agree that the way to beat this disease—as well as many other diseases—is through accurate, early detection.

There is a common precept in healthcare that says time equals information, which equals image quality. But with AI—specifically with the AIR Recon DL application—we can shorten the time it takes to take an MRI, and decrease the amount of image errors, while maintaining the quality of the image. Shorter and more accurate MRIs also allow the opportunity for more patients to be seen.

More than a product

Since it’s clear that AI has the potential to impact countless processes across healthcare, it’s important to note AI itself is not a product. The purpose of AI is to create better and more efficient processes. Across the world, healthcare systems are being asked to do more with less, unlike ever before. Efficiency is the key not just to patient health but also the well-being of clinicians.

Clinicians spend as little as 27% of their time with their patients[2]—the rest of their time is  spent poring over data, tracking down results, and doing cumbersome administrative work. That’s not why clinicians went into healthcare, and that’s not where they make their biggest impact. By using AI to improve processes, we can put clinicians back in front of the patients delivering care.

Simplifying and streamlining

For example, the Sonos CNS ultrasound application shows how AI built into the device can help make the operation easier and more accurate. During a 20-week check-up for pregnant women,  the central nervous system of the fetus is typically checked. In that monitoring process, the technician can perform up to 41 keystrokes but, with a deep learning model built into the device, the keystrokes can be reduced to as little as eight. That helps streamline the process and removes opportunity for error.

Also, in an effort to improve consistency in breast cancer screenings, the Breast Assist, by Koios[3] application uses an algorithm based on 400,000 images to provide a color-coded confidence indicator of the suspicion of breast lesion directly in the imaging system.

The journey is just beginning

We’re still at the start of this AI revolution, but it’s here to stay. 80% of healthcare leaders say they are investing or have invested in AI[4]. There are also more than 200 start-ups in healthcare[5] looking to create applications that meet the unique needs of clinicians all over the healthcare landscape. And GE Healthcare is right there at the forefront, providing a developer platform—including our Edison™Open AI Orchestrator—that allows developers across the world to provide solutions that can be seamlessly integrated into GE Healthcare systems and products.

As AI helps us do more with less, gives us better information, and improves the patient and clinician experience, we’re well on our way to achieving precision health.

Karley Yoder, the Vice President and General Manager of AI at GE Healthcare, is speaking on how to make healthcare more intelligent at the CHIME20 virtual event to discuss these and many other projected advancements that AI can help drive across healthcare.

But she’s just one of many industry experts from GE Healthcare who is available to speak at upcoming events. Learn more about our speakers bureau and book one of us for your next event!


[2] Harvard  Business Review,  November  2018

[3] Not CE marked. Cannot be placed on the market or put into service until it has been made to complete with the medical device directive.

[4] MIT Technology Review Insights Oct 2019

[5] Healthcare Dive Jan 2020