Structural connectivity based on diffusion Kurtosis imaging

• Autor(a) principal: Loução, Ricardo Sérgio Gomes Almeida
• Data de Publicação: 2015
• Tipo de documento: Dissertação
• Idioma: eng
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: http://hdl.handle.net/10362/16097
• topic: Diffusion tensor imaging; Diffusion Kurtosis Imaging; Tractography; Structural connectivity; Domínio/Área Científica::Engenharia e Tecnologia::Engenharia Médica

Structural connectivity models based on Diffusion Tensor Imaging (DTI) are strongly affected by the technique’s inability to resolve crossing fibres, either intra- or inter-hemispherical connections. Several models have been proposed to address this issue, including an algorithm aiming to resolve crossing fibres which is based on Diffusion Kurtosis Imaging (DKI). This technique is clinically feasible, even when multi-band acquisitions are not available, and compatible with multi-shell acquisition schemes. DKI is an extension of DTI enabling the estimation of diffusion tensor and diffusion kurtosis metrics. In this study we compare the performance of DKI and DTI in performing structural brain connectivity. Six healthy subjects were recruited, aged between 25 and 35 (three females). The MRI experiments were performed using a 3T Siemens Trio with a 32-channel head coil. The scans included a T1-weighted sequence (1mm3), and a DWI with b-values 0, 1000 and 2000 s:mm2. For each b-value, 64 equally spaced gradient directions were sampled. For DTI fitting only images with b-value of 0 and 1000 s:mm2 were considered, whereas for the DKI fitting, the whole cohort of images were considered. To fit both DTI and DKI tensors, extract the metrics and perform tract reconstructions, the toolbox DKIu was used, and the structural connectivity analysis was accomplished using the MIBCA toolbox. Tractography results revealed, as expected, that DKI-based tractography models can resolve crossing fibres within the same voxel, which posed a limitation to the DTI-based tractography models. Structural connectivity analysis showed DKI-based networks’ ability to establish both more inter-hemisphere and intra-hemisphere connections, when compared to DTI-based networks. This may be a direct consequence of the inability to resolve crossing fibres when using the DTI model. The DKI model ability to resolve crossing fibres may provide increased sensitivity to both inter- and intra-hemispherical connections. DTI-based modularity connectograms show a distinct intra-hemispherical configuration, whereas DKI-based connectograms show an increased number of inter-hemispherical connections, with several clusters extending over both hemispheres. Global and local connectivity metrics were also studied, but yielded no conclusive results. This may be due to a lack of reproducibility of the metrics or of the small cohort of subjects considered. DKI seems to provide additional insights into structural brain connectivity by resolving crossing fibres, otherwise undetected by DTI.