Real-time contextual information could help doctors interpret children’s brain scans

Radiologists who can tune in to the nuances of brain scans in children are a pretty rarified group. Only about 3 percent of U.S. radiologists, some 800 to 900 physicians, practice in pediatrics. Those specifically trained in pediatric neuroradiology are even scarcer.

To a less trained eye, normal developmental changes in a child’s brain may be misinterpreted as abnormal on MRI. Conversely, a complex brain disorder can sometimes appear normal. That’s especially true when the abnormality affects both sides of the brain equally (see sidebar).

It can be hard to find the cause of a child’s developmental delay without a proper read. “Pediatric brain scans of children under age 4 can be particularly tricky to read because the brain is rapidly developing during this period,” says Sanjay Prabhu, MBBS, a pediatric neuroradiologist at Boston Children’s Hospital. “If you’re looking at adult scans all the time, it’s incredibly difficult to transition to pediatric scans and understand what is considered ‘normal’ and ‘abnormal.’ Clinicians often wonder, ‘Should I repeat the scan? Should I send the patient to a specialist?’”

Decision support for brain scans in children

Prabhu and his colleagues are often asked to review brain scans taken at other hospitals. Frequently, they revise a prior diagnosis or make a first-time diagnosis for a child, ending a long and stressful diagnostic odyssey. Could they somehow share their expertise with a much wider group of physicians?

What is a normal brain? It depends on age

brain scans in children Boston Children's Hospital GE Healthcare

These MRI scans (click to enlarge) illustrate how providing reference images can improve the diagnosis of brain conditions in children. Many features of children’s brains are age-specific, such as the addition of the insulation known as myelin to nerve fibers. Myelination should occur symmetrically in both hemispheres of the brain, starting in the brainstem and ending in the frontal lobes.

The addition of myelin makes the white matter appear darker on the MRI scans. If you compare the two normal children, the 2-year-old’s white matter is much darker than the 6-week-old’s. This is true both in the deep regions of the brain (shown by arrows) and in the white matter of the cerebral hemispheres (asterisks).

The 2-year-old with developmental delay, due to Pelizaeus-Merzbacher disease, has a striking absence of myelination throughout the brain. Note the brightness at the posterior limb of the internal capsule (arrows) and in the white matter of the cerebral hemispheres (asterisks). This myelination pattern looks more like that of the normal 6-week-old than the normal 2-year-old. But because the pattern is symmetrical, it is easy to miss.

Today, Boston Children’s and GE Healthcare announced a collaboration to develop and commercialize digital diagnostic solutions. Their first project is a decision support platform aimed at improving diagnostic accuracy in pediatric brain MRI scans.

The platform will provide contextual information in real time, at the point of care, leveraging the software expertise of GE Healthcare, the high-volume computing power of the GE Health Cloud and the clinical know-how of radiologists at Boston Children’s. The system will come pre-loaded with normative reference scans from children of different ages. Doctors worldwide can use these as a benchmark when reading their own patients’ scans.

The hospital’s Innovation and Digital Health Accelerator (IDHA), led by Jean Mixer, VP Strategy and Digital Health, and John Brownstein, PhD, Chief Innovation Officer, also supports the project.

“This new digital tool, once available, will provide non-specialists with access to knowledge and expertise to help effectively diagnose children,” says Richard Robertson, MD, Boston Children’s radiologist-in-chief. “It will fill a large unmet need in medicine for pediatric radiology.”

Read more on this collaboration in the Boston Globe and in GE Healthcare’s newsroom.