Get access

Ultrafast Mr Imaging in Pediatric Neuroradiology


*Paul D. Griffiths, Section of Academic Radiology, Floor C, Royal Hallamshire Hospital, Glossop Road, Sheffield, S10 2JF, UK.


Purpose:  To compare the diagnostic information obtained from ultrafast MR imaging with standard MR imaging techniques in pediatric neuroradiology. The goal was to judge whether ultrafast methods can be used to replace standard methods and reduce the need for sedation or general anesthesia as a result of the considerably shorter scan times.

Material and Methods:  Our prospective study involved 125 patients. Routine clinical imaging was performed along with two ultrafast methods. Single shot fast spin echo (SSFSE) was used to give T2-weighted images and an echo planar imaging (EPI) sequence to provide a T1-weighted images.

The ultrafast images were presented to an experienced neuroradiologist who was also given the information present on the initial referral card. These reports based on the ultrafast images were then compared with the formal radiologic report made solely on the basis of the standard imaging.

Results:  The overall sensitivity and specificity for ultrafast imaging when compared to the reference standard were 78% and 98% with positive and negative predictive values of 98% and 76%. Pathologies characterized by small areas of subtle T2 prolongation were difficult or impossible to see on the ultrafast images but otherwise they provided reliable information.

Conclusions:  This paper demonstrates that ultrafast MR imaging can diagnose many pediatric intracranial abnormalities as well as standard methods. Anatomic resolution limits its capacity to define subtle developmental anomalies and contrast resolution limitations of the ultrafast methods reduce the detection of pathology characterized by subtle T2 prolongation.