The Critical Care system will pave the way for many more AI embedded (edge inference) and powered X-ray devices to enable more accurate and timely diagnosis for critical conditions, relieving the burden radiologists are facing with increasing patient caseloads.
On the 12th September 2019, GE Healthcare made history after it announced 510(k) clearance from the Food and Drug Administration (FDA) of the Critical Care Suite. The Critical Care Suite was built in collaboration with UC San Francisco (UCSF) and uses GE Healthcare’s Edison platform. The Critical Care Suite is first of its kind in the industry having artificial intelligence (AI) embedded algorithms on the mobile X-ray device, to scan X-ray images and more accurately detect the deadly condition pneumothorax. Pneumothorax, also referred to as a collapsed lung, affects approximately 74,000 Americans a year.
The AI powered mobile X-ray device enables the time required for radiologists to review a suspected pneumothorax to reduce significantly, from eight hours to as little as 15 minutes, and enables prioritization of most critical cases. When a patient is scanned using the Critical Care Suite, images are automatically examined by a simultaneous search for pneumothorax. In cases where there is a suspected case of pneumothorax, the radiologist is sent an alert alongside the original chest X-ray for automatic review by a Picture Archiving and Communication Systems (PACS). In addition to this, the technologist performing the scan will also receive a notification for the case to be prioritized and to raise awareness of the potential diagnosis. Other AI algorithms used in the scanning procedure, which were developed using GE Healthcare’s Edison platform, include quality focused AI algorithms, which highlight at the point of scan, potential field of view errors and protocol. These quality checks on the technologist work flow enable point of action to be made by the technologist to ensure rejections or reprocessing are rectified before the image is sent to the PACS.
The Critical Care Suite eliminates any reliance on transfer speeds or connectivity to implement AI aided diagnosis, as AI results are generated within seconds of image acquisition, enabling speed and reliability and disregarding any processing delay. Therefore, the functions of the technology are not reliant on an internet connection or the cloud, allowing hospitals the opportunity to use the AI algorithms without additional investment into IT infrastructure.
IHS Markit perspective
The Critical Care system will pave the way for many more AI embedded (edge inference) and powered X-ray devices to enable more accurate and timely diagnosis for critical conditions, relieving the burden radiologists are facing with increasing patient caseloads. A combination of radiologist burnout, technician error and increasing patient footfall are just some of the current issues hospitals are facing universally, which systems like the Critical Care Solution can help to address. Now that the seal has been broken, GE Healthcare, alongside its rival medical imaging leaders, will continue working toward integration of AI into every part of the healthcare system to improve inefficiencies, improve clinical outcome of the patient and eradicate errors. Among the top early use cases and potential benefits for AI in healthcare include:
- Alleviating repetitive, high-volume tasks (cancer detection and screening).
- Reducing imaging time and radiation dose.
- Providing triage and prioritizing urgent cases, such as the case with GE’s latest approval for the Critical Care Suite (the pneumothorax algorithm embedded on Optima XR240amx).
- Providing automatic quantification to replace manual measurements.
- Facilitating non-invasive imaging techniques.
- Facilitating faster detection of rare disease.
- Enabling examination of temporal relationships of pathological findings on images in context of other information.
In its recent research, IHS Markit has reviewed the lung-related AI offerings from more than 35 companies, most of which are focused on TB screening (AGFA Healthcare) and lung cancer screening (4Quant, Enlitic, HealthMyne, Infervision, Lpixel, Samsung, Siemens, TeraRecon, Thirona, and Zebra Medical Vision among others), but few have FDA approval. The only other example of embedded AI is Samsung’s GC85A. As such, this announcement from GE Healthcare is notable as an example of edge inference with regulatory approval, and is an example of a use case that aligns with provider needs and workflow.
The global general radiography market is valued by IHS Markit to reach $2,726.2 in 2023, with the digital X-ray market, including both fixed and mobile digital X-ray systems, expected to account for 78% of the total market. The clinical diversity and operational efficiency mobile X-rays offer are factors which will fuel the mobile X-ray market to grow at a CAGR of 7.4% for global units through to 2023. In developed regions where the installed base of both fixed and mobile X-ray systems is currently wholly digital, replacement to the next generation of digital X-ray systems is expected to be AI powered and algorithm embedded such as the Critical Care System. In a current healthcare environment, group purchasing organizations and administration departments are increasingly taking control of purchasing decisions for hospital facilities. Such AI developments embedded in the X-ray systems themselves will be crucial in justifying additional investment, to reap the benefits of increased clinical precision, extensive time saving measures and increased return of investment.