Faster, safer pediatric MRI scans: MCHRI Faculty Scholar transforms diagnostic imaging for children

June 2, 2021

By Roxanna Van Norman

Magnetic resonance imaging (MRI) scans are notoriously loud, take a long time to complete, and require the patient to lay still. Even though MRI technology is widely used in hospital and clinics and extraordinarily helpful in providing detailed, high-resolution to detect problems in the body, the overall experience is tough on kids and often yield blurry images.

"The challenge for us in pediatric imaging is that children do not cooperate as well as adults do during an MRI scan, which makes the imaging difficult," says Shreyas Vasanawala, MD, PhD, Professor of Radiology at Stanford University School of Medicine (SoM). 

If a child moves during the exam the images become blurry – similar to taking a still photo of a fast-moving car; the images will come out hazy, obscuring pertinent information. Healthcare providers often use general anesthesia or opt for alternative diagnostic imaging involving radiation to get an accurate scan of a child's body.

Wanting to make MRI scans easier for children, Dr. Vasanawala proposed a plan to transform the commonly performed procedure by using rapid MRI methods designed to making imaging faster, safer, and readily available for use in hospital settings. The Stanford Maternal and Child Health Research Institute (MCHRI) named him the Tashia and John Morgridge Faculty Scholar in Pediatric Translational Medicine (2009-2017), which came with total funding of $430,000 to support his idea.

"If you can speed up the imaging, then you can make it more child-friendly. Further, with new algorithms to detect the motion and correct for it, you can still have sharp images even when there is motion," says Dr. Vasanawala. With support from MCHRI, he focused on developing rapid MRI methods for evaluating congenital heart disease in children. 

Building a multi-disciplinary team

Dr. Vasanawala is the Division Chief of Pediatric Radiology in the Department of Radiology at SoM and the Radiologist-in-Chief of Stanford Children’s Health. He has spent the last 12 years studying how to improve MRI scans for children.

A full cardiac MRI exam for a child with congenital heart disease involves capturing three components to render an overall picture of the child's heart. The first component obtains four-dimensional anatomical images of the organ – three spatial dimensions along with the fourth dimension of time as the heart is beating. The second component quantifies the flow of blood through various vessels and heart valves. The last part measures the heart's function - how well the muscle of the heart is contracting with each heartbeat.

“Altogether, these three components historically took 90 minutes to get a comprehensive MRI scan," says Dr. Vasanawala. During this lengthy process, the patient would have to remain motionless the entire time and temporarily hold their breaths. He approached conventional MRI scans with a fresh perspective – could an MRI scan collect minimal data and still reconstruct the overall image and deliver the same results?

Dr. Vasanawala teamed up with scientists and researchers across various disciplines to address the challenge of dramatically speeding up MRI scanning and removing the need for anesthesia in children. He comprised a multi-disciplinary team of industry experts at General Electric Healthcare and faculty within the Departments of Computer Science, Electrical Engineering, and Radiology at Stanford. Co-investigators included John Pauly, PhD, Professor of Engineering and Co-Director of the Magnetic Resonance Systems Research Laboratory; Brian Hargraves, PhD, Professor of Radiology; and former Stanford graduate student and postdoctoral fellow Michael Lustig, MSc, PhD, now Associate Professor of Electrical Engineering and Computer Science at UC Berkeley. 

Together, the researchers worked closely with personnel at Lucile Packard Children’s Hospital (LPCH) Stanford, including many nurses and MRI technologists.

Innovating the conventional MRI scan

A major limitation of a traditional MRI scan is its slow imaging speed. Dr. Vasanawala and team’s work has helped to reduced cardiac MRI time for some exams from 90 minutes to 10 minutes. 

Early in the project, Dr. Vasanawala conducted a study looking at pediatric patients referred to LPCH for cardiovascular, abdominal, and knee MRI scans. The scans integrated a novel imaging technique known as compressed sensing to enable researchers to construct images from fewer scan measurements while maintaining high imaging accuracy.

The team’s research combined traditional imaging methodologies with compressed imaging, which enabled faster scans, higher image resolution, and greater potential for use in a pediatric clinical setting as compared to traditional MRI. Another study involving pediatric patients with congenital heart disease led to similar results.

"We've done work where we're able to form images from very little measurements, so the system has to learn how to reconstruct images from a short scan with very little data," says Dr. Vasanawala.  

Another significant accomplishment involved a study to assess the use of iron-enhanced MRIs in babies for complex congenital heart disease – all without the use of anesthesia. Findings demonstrated a higher diagnostic performance and detailed assessment of the heart valves and the coronary arteries.  

With a growing body of research, Dr. Vasanawala's group has been recognized worldwide for their clinical and translational impact in pediatric imaging methods, including expanding the use of pediatric-specific MRI receiver coils to hospitals and clinics.

Expanding the research portfolio

Funding from MCHRI provided critical support for Dr. Vasanawala and his research team to kick start a proposal to the National Institutes of Health (NIH) for anesthesia-free pediatric MRI. Still, without a compelling track record to support the high-impact study in pediatric diagnostic imaging, getting funding was difficult. 

Dr. Vasanawala applied to the NIH three times and was rejected the first two. On the third time, the National Institute of Biomedical Imaging and Bioengineering awarded him a Research Project Grant (R01) in 2010 to support this work to develop a rapid, robust pediatric MRI. The team also received competing grant renewals in 2014 and 2018.

"That initial belief in us, a little bit of faith, was critically helpful," says Dr. Vasanawala. "Without MCHRI support, I wouldn't have had a way to obtain the preliminary data [to help support the application to the NIH]." 

Dr. Vasanawala also recalls the challenges in navigating the research landscape as an early-career investigator, including roadblocks in addressing regulatory approvals for the study. Recognizing the need for this clinical research support and regulatory guidance, MCHRI established the Clinical Research Support Office (CRSO) to provide resources and guidance for investigators to navigate the complex process of conducting studies at SCH. 

Now, more than a decade later, Dr. Vasanawala serves as the Radiologist-in-Chief for SCH where multiple radiology faculty are developing innovative diagnostic and therapeutic options for children. The team has built an expansive research portfolio in developing faster MRI scan techniques, advancing novel applications for obtaining high-resolution images with much less anesthesia, and employing these transformative imaging methods in hospital and clinical settings.

"Our work has expanded in many directions of pediatric vascular, oncologic, gastrointestinal, liver, and kidney diseases, and it was catalyzed by that initial MCHRI support that kicked it all off," he says.

Roxanna Van Norman is the marketing manager for the Stanford Maternal and Child Health Research Institute.