A Broader Platform
By Keith Loria
Radiology Today
Vol. 21 No. 4 P. 10

Advances in MR improve speed, quality, and patient experience.

There are some facets of MR equipment that manufacturers will always look to improve—finding faster scans, getting better images, improving workflow and productivity for the technologist, and making things more comfortable for the patient. The innovations presented at RSNA 2019, as well as the advancements that companies have been touting for the past couple of years, indicate that all of these areas are improving. And with AI starting to make more of an impact on MR capabilities, further improvements seem likely in the years ahead.

More With Less
Michael Brandt, chief marketing officer of global MR for GE Healthcare, says the company is working on making scans faster, making systems more personalized, and helping customers use quantitative data more effectively. “It turns out, if you make the patient more comfortable and the scans better, it improves the patient experience a great deal,” Brandt says.

Brandt confirms that the rise of AI and analytics in MR is generating the biggest buzz in 2020. RSNA 2019 even had a separate hall for the technology, with health leaders sharing information and learning about what is being done. “While AI has opportunities in a lot of modalities, I think it has some serious opportunities in MR,” he says. He explains that, because clinicians are expected to read through an increasing volume of data, AI and analytics can help manage the volume as well as assist in postprocessing efforts. The newest AI application from GE in MR is AIR Recon DL (deep learning), which is 510(k) pending.

“It’s a reconstruction technology that works from the raw data and is built on our Edison platform,” Brandt says. “It’s a deep learning, neural network–learned tool designed to improve the signal-to-noise extraction from the raw data. This will allow the clinician to make a resolution decision or a scan performance decision and improve image quality. With some compromises, you could even conceivably get both.”

Tom Naypauer, senior manager of MR solutions marketing for Canon Medical Systems USA, says the emergence of AI in medical imaging is the biggest trend today and its impact on MR has been a hot topic for the past couple of years. “This past RSNA, we started to see AI products introduced. From a Canon Medical perspective, we introduced AiCE (Advanced intelligent Clear-IQ Engine) deep learning reconstruction on the Galan 3T MRI and the Orian 1.5T MRI,” Naypauer says, “For the first time, we are simultaneously able to improve signal to noise and resolution without negatively impacting speed.”

He explains that, in the past, improving one of those three parts of the triangle meant giving up something. AiCE is designed to provide exceptional image quality, specifically a high signal-to-noise ratio, and resolution that was not clinically practical previously.

Arjen Radder, PhD, business leader of global MRI at Philips, says radiologists today want to do more with less, and that concept has been a big factor in the MR equipment being released lately. “That is what’s driving productivity,” Radder says. “It’s all about the experience of patients but also of health care staff. In MR, the needs of the customers are very much built around this concept, also summarized as the Quadruple Aim.”

MR typically has long scanning times, and Radder notes the entire industry is looking at ways to shorten them and do more MR per time unit. Additionally, arriving at a diagnosis quicker is also a problem that all manufacturers are looking to solve, and Philips has worked to reduce scan time. An example of this is the company’s Compressed SENSE, an acceleration technique that shortens single MRI sequences and full MRI examinations. Radder says this allows radiologists to choose faster scanning times, more time spent with patients, or more data collection per patient without additional scan time.

“We have more than 3 million patients who have been scanned with Compressed SENSE,” Radder says. “They have averaged a reduction in scan time of 30% to 50%, which means more scans can be done per time unit, which is a productivity gain. Some radiologists use the additional time to get more information for each patient to come to a more definitive diagnosis.” Radiologists can also choose to have fewer working hours, which can reduce overtime as well as stress levels.

“AI and clinical decision support are playing an important role there,” Radder says. “When it comes to the experience—both for the patient and staff—it’s all about making sure we understand workflow and overall experience, and really trying to improve in those areas.”

Patient Contemplation
GE’s AIR family of products has three pillars—coils, workflow, and image quality. Together, the trio is designed to deliver clinical versatility, better patient comfort, and intelligent productivity improvement. Brandt says it also provides consistent and improving image quality.

The company is rolling out new AIR coils that allow technologists to scan patients more easily and position them more comfortably. For example, the lumbar spine can be imaged by having an AIR coil draped over the patient without the need for reactivating the posterior coil. “The coil also allows wrapping completely around the foot, so you can scan the whole foot, including the sole, which in the past has been more challenging,” Brandt says.

Another example is a patient who needs to return for multiple scans, Brandt says. The company’s AIRx product is a slide positioning tool that will, irrespective of the patient’s position, put the slices in the same plane. This means less read-through required by the radiologist and minimizes the time the patient is in the machine.

About two years ago, Siemens Healthineers North America introduced MAGNETOM Vida, a 3 T MRI that was the first system to feature BioMatrix technology. From there, Siemens Healthineers rolled out MAGNETOM Sola, a 1.5 T system and, more recently, two newer systems: MAGNETOM Altea 1.5 T and MAGNETOM Lumina 3 T. Chris Farischon, MR product manager at Siemens Healthineers, explains how the Altea and Lumina utilize it to adapt to patients’ unique traits and variabilities.

“Typically, in the past, we relied on patients to remain still during an exam while also holding their breath,” Farischon explains. “BioMatrix technology accounts for patient movement, which gives the technologist more flexibility so they don’t have to start over or rescan.”

Faster Workflow
An additional workflow benefit is the Select&Go interface, which is common across the company’s MR platform and helps speed up exam time by positioning the patient at isocenter faster. Farischon explains that Select&Go is a touchscreen that sits on the side of the magnet and resembles an iPad. Once a patient lies on the table and after the technologist plugs in the coil and selects the anatomy to be scanned, the AI-based system will automatically move the patient to isocenter based on traits such as the patient’s height and the coil being used. “The other unique benefit is that it prefetches the exams that are related to that particular coil,” Farischon says. Another technology that Siemens employs is Dot engines, which simplify MR exams by speeding up workflow and allowing introductory technologists to conduct more advanced exams, eg, cardiac imaging, he notes.

Philips has made workflow improvements with its VitalEye platform, which enables touchless sensing to detect patient physiology and breathing movement during image acquisition. When used in combination with VitalScreen, routine exam set-up time is less than a minute, regardless of operators’ experience. Radder notes that these tools help to standardize and speed up workflow. “The goal is to have more measurements done during scanning, and that helps with correction for image quality,” he says. To help meet that goal, Philips has automated a number of protocols on its MR equipment, allowing technologists to push one button, leave the room, and begin the scan.

Naypauer says Canon has a few ways of making things easier for radiologists and clinicians. “For example, CardioLine+ automatically generates results in up to 14 different planes for MR,” he says. This includes standard cardiac planes in the right and left ventricles as well as the four cardiac valves in a single breath-hold scan. “From a radiologist perspective, this gives consistent results in terms of eliminating the variability of different users planing on the systems.” Additionally, it’s useful for comparing scans over time, he says. At RSNA 2019, Canon Medical also introduced its Compressed SPEEDER MR technology, which Naypauer explains makes acquisition four times faster while maintaining image resolution.

More Work Ahead
Hospitals and health care clinics that want to take advantage of the latest MR technology need to stay up to date with new platforms and hardware. “It’s quite clear that in acceleration there’s more to do beyond Compressed SENSE, and that’s where new developments in AI come in,” Radder says. Philips also recently introduced its helium-free-operations Ambition platform, which reduced the system’s helium requirement from 1,500 to 7 liters, resulting in a smaller, lighter, and more flexible installation footprint. Radder notes that helium costs have risen significantly in the past couple of years.

Naypauer says AI is still in its infancy and will drastically change MR as the technology becomes more advanced. “We are starting to see these advanced AIs applied more routinely,” he says. “We have applied it in the areas of image quality and reconstruction. One other area where I think AI would make a big impact is in workflow and productivity. Automating mundane tasks that technologists have to perform every day with every patient could have huge potential for productivity improvements as well.”

Brandt expects increasingly faster scanning in upcoming years and tools that prepopulate parts of a report. He also foresees technology that compensates for and negates patient motion. However, he thinks these innovations will take time.

“The discussion piece is easier than actually doing it, and often times we trivialize what it actually takes to do this,” Brandt says. “As solutions for these products come out, we may hit delays in getting to market as fast as we’d want.”

— Keith Loria is a freelance writer based in Oakton, Virginia. He is a frequent contributor to Radiology Today.