GE Healthcare drew keen attention by unveiling Edison, an intelligence platform that connects imaging equipment with data to collect and manage the data using artificial intelligence technology, at the meeting of the Radiological Society of North America last year.
Based on Edison, which can combine AI technology and cloud connectivity to process data quickly, GE Healthcare released new devices and applications, including CT, X-ray, MR, and ultrasound equipment.
The latest move of GE Healthcare showcased that the use of AI is not an option but a necessity in the healthcare industry.
What does GE Healthcare seek through AI, then? Korea Biomedical Review recently met with GE Healthcare Chief Medical Officer Mathias Goyen and Chief Commercial Officer Scott Miller to hear their answers to this question. They visited Korea to attend the meeting of the Korean Society of Radiology (KCR) 2019.
|Scott Miller (left), Chief Commercial Officer at GE Healthcare, and Mathias Goyen, Chief Medical Officer, in an interview with Korea Biomedical Review
KBR: How do you facilitate AI in product development?
Goyen: A huge amount of medical data is being generated every day, and we are paying attention to AI as a key to advancing medicine using such data. AI is applied not only in the scan stage but also before and after the scan. The integration of AI in these three phases is the key to GE Healthcare’s use of AI. I’m not aware of any start-up that uses AI in all stages – pre-scanning, execution, and post-scanning.
KBR: Can you elaborate on the three stages of AI use?
Goyen: For example, when doing MRI, the radiologist starts by positioning the patient’s head. Because the head size of each patient is different, the position of the head of the patient in the bore of the scanner. As the anatomy of each patient is different, the position of the head has to be finely adjusted to obtain the desired image. However, GE Healthcare's MRI combined with AI captures the patient’s anatomical positions and automatically recognizes its locations, positioning the brain where it is needed and automatically acquiring images. Even if the patient takes an MRI several times, the images are taken from the same location, which enables more accurate analysis.
Just like this, AI is applied from the pre-test stage. I can take an example of an ultrasound scan as the test stage. With the help of built-in AI capabilities the radiologist can easily differentiate between arteries and veins due to the different flow pattern by initiating the algorithm after having placed the transducer in the region of interest. Also, AI-powered X-rays help medical experts diagnose pneumothorax more quickly and accurately.
AI not only provides benefit for imaging devices but helps improve hospital systems. A German medical group, which has nine affiliated hospitals, introduced MRI equipment with GE Healthcare's Edison platform. It shortened patients’ MRI schedule from six weeks to two weeks.
KBR: Many companies launch AI-based products, but they do not talk about the quality of the AI technology they’re using. How does GE Healthcare manage the quality of AI?
Goyen: We have developed AI by combining our excellent talents with outside partners. The AI technology to detect pneumothorax was the collaboration work of four hospitals, including the University of California San Francisco Hospital.
Miller: What GE Healthcare wants is to work with safe, transparent and efficient partners. AI development requires a variety of rich data. We need good “input” to develop a robust algorithm. So, the availability of diverse and accurate data is an important factor in selecting a partner. We run rigorous testing to verify secured data.
KBR: Do you have any plan to use Korean data?
Miller: AI technology needs abundant and various data. Research in radiology is quite advanced in Korea, and there are many studies. As Korea is leading in terms of AI operation, there is a possibility that we may use Korean data.
KBR: Do you think AI will reduce the gap in physicians’ skills?
Miller: Yes. AI will bring disruptive medical innovation. Indeed, as AI is applied to devices and cloud, it is creating collaborations that are different from existing ones. Physicians’ gaps are being narrowed through these platforms.
KBR: As AI drew attention, some say AI may replace doctors. What do you think?
Goyen: AI is not science fiction but a scientific fact. There are many repetitive, easy but stale tasks in the medical field. For example, radiologist should find and check dozens of metastases. But AI can do this kind of work faster and more accurately. However, AI per se will not replace radiologists. Rather, radiologists who leverage the power of AI may replace those who don’t. It only redirects more time for the doctor to see the patient better. I hope doctors do not be afraid of AI but embrace it.
<© Korea Biomedical Review, All rights reserved.>