Diagnostic performance of diffusion-weighted MR neurography as an adjunct to conventional MRI for the assessment of brachial plexus pathology.

objectiveTo investigate the diagnostic performance of diffusion-weighted (DW) MR neurography as an adjunct to conventional MRI for the assessment of brachial plexus pathology.

methodsDW MR neurography scans (short tau inversion recovery fat suppression and b-value of 800 s/mm

resultsMedian interpretation times ranged between 20 and 30 s. Interobserver agreement was substantial (κ coefficients of 0.715-0.739). For the 5th year radiology resident, sensitivity was 53.3% (95% CI, 30.1-75.2%) and specificity was 100% (95% CI, 92.1-100%). For the junior neuroradiologist, sensitivity was 66.7% (95% CI, 41.7-84.8%) and specificity was 100% (95% CI, 92.1-100%). For the senior neuroradiologist, sensitivity was 73.3% (95% CI, 48.1-89.1%) and specificity was 95.6% (95% CI, 85.2-98.8%). Traumatic injury, metastases, radiation-induced plexopathy, schwannoma, and inflammatory process of unknown cause could be detected by the majority of readers (100% detection rate for each disease entity by at least two readers). Neuralgic amyotrophy, iatrogenic injury after first rib resection, and cervical disc herniation causing root compression were not detected by the majority of readers (0% detection rate for each disease entity by at least two readers).

conclusionDW MR neurography may be a useful adjunct when assessing for brachial plexus abnormalities, because interpretation time is relatively short and the majority of abnormalities can be detected. KEY POINTS: • DW MR neurography interpretation time of the brachial plexus is relatively short (median interpretation times of 20 to 30 s). • Interobserver agreement between three readers with different levels of experience is substantial (κ coefficients of 0.715 to 0.739). • DW MR neurography can detect brachial plexus abnormalities with moderate sensitivity (53.3 to 73.3%) and high specificity (95.6 to 100%).