This study aims to define lesion localization patterns and their relationship with clinical features in patients with pNBD using probability mapping analysis.

Behçet’s disease (BD) is a multisystemic inflammatory disorder. The most common neurological involvement of BD is parenchymal neuro-Behçet’s disease (pNBD) that mainly affects brainstem and diencephalic structures.

We included 66 non-standardized MRI’s of 55 patients with pNBD. T2-weighted axial images were digitalized using computer-aided design software. Lesion boundaries were determined as polygons and converted into high definition raster datasets. Later, digitalized images were transferred into the brain template, and spatial analyses were performed to sum up intersected boundaries. Lesion probability maps were obtained according to clinical and laboratory subgroups. Finally, subtraction maps were created to compare the subgroups.

A total of 66 MRI’s of 55 patients were used to generate the lesion probability maps. Based on the lesion topography, most affected regions were pontomesencephalic junction, pons, and corticospinal tracts. Brainstem lesions were more common in females, whereas basal ganglia lesions were more common in males (p<0,01). In progressive and more severe disease, corticospinal tracts were affected more frequently (p<0,01). The presence of uveitis and a positive pathergy test was associated with lesions in the mesodiencephalic junction (p<0,01).

To date, parenchymal lesions in NBD were analyzed using rough anatomical regions in prior studies. In this study, we employed a novel technique to create lesion probability maps using unstandardized MRI’s. Therefore, our study provides accurate lesion localizations with objective visual maps in various subgroups of NBD. A further study with more patients will be able to define the relationship between lesion localization and the prognosis of pNBD.