3T MRI distinguishes signs of child abuse

April 21, 2006

Though an MR scanner is hardly cuddly, it could become a baby’s best friend. Or so Christopher Looney would say.

Putting an infant inside an MR scanner for brain imaging studies may be a daunting endeavor. Many mothers would be reluctant to let their offspring undergo such testing, and the babies themselves are hardly ever cooperative. Patience and the right setup - including dim lights and soothing music - could provide for rewarding results,according to radiology residency candidate Christopher Looney.

Though an MR scanner is hardly cuddly, it could become a baby's best friend. Or so Christopher Looney would say.

The 29-year-old University of North Carolina medical student just finished interview rounds with radiology departments across the country. His 2005 RSNA Student Research Trainee Prize may or may not help him land his dream job. But the $1000 check certainly helped him cover some traveling expenses.

He's not yet sure what subspecialty area he might choose in the long run. There is a strong chance, though, he will fall for neuroradiology. Here's why.

Intracranial hemorrhage in asymptomatic babies is supposedly an unusual finding. So clinicians thought, until a group of UNC investigators, including Looney, did a study that ended up challenging that assumption.

They had originally set out to study normal brain development in neonates using 3T MRI, but they unexpectedly found a significantly high prevalence of intracranial hemorrhage in the babies studied. Many questions arose. As a result, Looney and his colleagues delved more deeply into the causes of these bleeds.

"Several months into the study, Dr. J. Keith Smith, the consulting neuroradiologist for the group, started to see these bleeds," Looney said. "The babies in the study were completely healthy, normal ones. We thought that studying the prevalence of intracranial hemorrhage in healthy babies would be interesting and a valuable experience."

Smith, Looney, and colleagues had prospectively enrolled several pregnant women for their original work on prenatal and neonatal brain development. They included 79 babies one to five weeks old for the intracranial hemorrhage review and imaged them on a 3T scanner without sedation. They also checked their obstetric records to determine hemorrhage risk factors.

They confirmed a 10% intracranial hemorrhage prevalence in the babies imaged, along with other surprising findings. For one, vaginal birth was significantly associated with the bleedings. Neuro- and pediatric radiologists at UNC see these bleeds fairly often, but they were still surprised at their prevalence in the study, Looney said.

"It would be premature to suggest to pregnant women that C-sections should be done with more frequency. We need to see if these bleeds resolve and also if they have any clinical impact. Our sample population was rather large for a radiology study but not big enough to investigate any one particular risk factor in depth, such as the use of forceps or vacuum delivery," he said.

Another important finding was that these hemorrhages differ significantly from those associated with nonaccidental head trauma.

"Something that is almost as significant is for radiologists to be able to tell the difference between these bleeds and nonaccidental head injuries," Looney said.

Intracranial bleeding associated with abuse has a different imaging appearance. Hemorrhages in Looney's study were typically infratentorial or peritentortial and subdural. Hemorrhage associated with nonaccidental head injuries, on the other hand, is typically supertentorial and intrahemispheric.

The age of injury provides more clues. A baby with a nonaccidental brain hemorrhage could have several hemorrhagic spots detected on MR. But radiologists would be able to tell that these bleeds had occurred at different times because of the difference in hemoglobin decompensation, which varies heavily with time. Bleeds associated with vaginal births occur simultaneously.

Though these findings were interesting, they need further testing, Looney said. For now, he is concentrating on finding the best residency program.

"I'm looking for a residency program that has a strong academic research track, because I'm very interested in continuing research, and I plan to establish a career in academic radiology," he said.

Other things out there catch his eye. Advances in tumor ablation have him thinking of interventional radiology. Recent imaging technology breakthroughs beckon to body imaging. And his own experience as a high-school and college athlete put musculoskeletal radiology very high on Looney's radar. His neuroradiology experience with babies, however, may have struck a chord.

"The mind is one of the last frontiers. There's so much to be discovered and so much to be learned in neuroradiology. It's difficult to know if the place where I go will match my research interests, but I hope it will work out that way," Looney said.

Using 3T MRI, Looney and colleagues found an unusually high number of intracranial hemorrhages in newborn babies, most associated with vaginal birth. Figures 1 and 2 show infratentorial subdural hemorrhages in the axial and sagittal planes, respectively.