Neurite orientation dispersion and density imaging parameters may help for the evaluation of epileptogenic tubers in tuberous sclerosis complex patients.

Abstract

To investigate the usefulness of neurite orientation dispersion and density imaging (NODDI) in evaluating cortical tubers, especially epileptogenic tubers in tuberous sclerosis complex (TSC) patients. High-resolution conventional MRI and multi-shell diffusion-weighted imaging were performed in 27 TSC patients. Diffusion images were fitted to NODDI and DTI models. Tubers were visually assessed on different image types and scored by two neuroradiologists. For 10 patients who underwent epilepsy surgery, the contrast ratios between lesion and background tissue were measured on different image types, and these were compared between 16 epileptogenic tubers and 92 non-epileptogenic tubers. There were significant differences in lesion conspicuity scores and lesion-background contrast ratios across different sequences (both p< 0.001). The post hoc analysis showed that both the conspicuity scores and contrast ratios of intracellular volume fraction (ICVF) derived from NODDI were higher than other image types. For the 16 epileptogenic tubers, lesion visibility on ICVF was better/equal in 4/12 tubers compared with conventional MRI and better/equal in 5/11 tubers compared with DTI. Significant differences were observed between epileptogenic and non-epileptogenic tubers on diffusion maps, especially on orientation dispersion index derived from NODDI (p< 0.0001). ICVF demonstrated higher contrast than conventional MRI and DTI, which helped detection of subtle epileptogenic tubers. Moreover, NODDI parameters showed the potential to identify epileptogenicity. • The noninvasive localization of epileptogenic cortical tubers is essential for the preparation of epilepsy surgery for TSC patients. • ICVF derived from NODDI showed greater contrast than conventional MRI and DTI in detecting tubers, especially subtle epileptogenic ones. • Diffusion parameters, especially ODI derived from NODDI, can support the identification of epileptogenicity.