Human cervical spinal cord funiculi: Investigation with magnetic resonance diffusion tensor imaging

Abstract

Purpose: To use spinal cord diffusion tensor imaging (DTI) for investigating human cervical funiculi, acquire axial diffusion magnetic resonance imaging (MRI) data with an in-plane resolution sufficient to delineate subquadrants within the spinal cord, obtain corresponding DTI metrics, and assess potential regional differences. Materials and Methods:Healthy volunteers were studied with a 3 T Siemens Trio MRI scanner. DTI data were acquired using a single-shot spin echo EPI sequence. The spatial resolution allowed for the delineation of regions of interest (ROIs) in the ventral, dorsal, and lateral spinal cord funiculi. ROI-based and tractography-based analyses were performed.Results: Significant fractional anisotropy (FA) differences were found between ROIs in the dorsal and ventral funiculi (P ¼ 0.0001), dorsal and lateral funiculi (P ¼ 0.015), and lateral and ventral funiculi (P ¼ 0.0002). Transverse diffusivity was significantly different between ROIs in the ventral and dorsal funiculi (P ¼ 0.003) and the ventral and lateral funiculi (P ¼ 0.004). Tractography-based quantifications revealed DTI parameter regional differences that were generally consistent with the ROI-based analysis. IN THE LAST TWO DECADES basic research on spinal cord injury and pathology has made exciting progress in identifying factors that beneficially influence spinal cord repair and regeneration. In the human, the evaluation of treatment efficacy and spinal cord fiber tract integrity in vivo is limited to behavioral sensorimotor assessments (1-4). Notwithstanding the importance of these assessments there is a growing interest to noninvasively assess human spinal cord white matter micro-architecture in vivo (5-9).Diffusion tensor imaging (DTI) is a noninvasive magnetic resonance technique that is sensitive to the diffusion of water molecules within and across extracellular spaces of tissues. Anisotropic diffusion in the central nervous system (CNS) is caused by barriers to diffusion, such as cell membranes and myelin found in CNS white matter (WM). Higher anisotropy in WM compared to gray matter (GM) has been demonstrated in vivo in both animals and humans and it is generally understood that the primary reason for this higher anisotropy in WM is due to packed axon bundles in various WM regions (10,11). Although WM delineation with DTI is well established in brain, the application of DTI in the spinal cord is more difficult because of its small cross-sectional size, susceptibility artifacts, and motion artifacts from cerebrospinal fluid (CSF) pulsation and respiration.The diameter of the human cervical spinal cord enlargement is %13 mm. In cross-section the spinal cord is typically characterized by the butterfly-shaped GM surrounded by WM. The WM on each side of the spinal cord can be grossly divided into dorsal, lateral, and ventral funiculi. The dorsal funiculus is the spinal cord WM on either side between the dorsal median sulcus and the dorsal root. Similarly, the ventral funiculus is the WM lying on either side between the ventral me...

Age-related changes of the diffusion tensor imaging parameters of the normal cervical spinal cord

“…We observed that the FA and ADC values indicated significant differences between GM and individual WM funiculi. These data are supported by previous studies [1,21]. In addition, we found that there were significant differences in the FA and ADC values of the different WM funiculi at different CSC segments, but that there were no differences in the DTI parameters of the GM from different CSC segments, an analysis that was not performed by Vedantam et al Compared with the upper and middle segments, both FA and ADC values from WM funiculis of lower segments were lower.…”

“…In addition, we found that there were significant differences in the FA and ADC values of the different WM funiculi at different CSC segments, but that there were no differences in the DTI parameters of the GM from different CSC segments, an analysis that was not performed by Vedantam et al Compared with the upper and middle segments, both FA and ADC values from WM funiculis of lower segments were lower. These results could be due to the decreased spinal cord volume of the lower segments, or to differences in signal-to-noise ratio (SNR) across the cervical segments [1,11,21].…”

“…However, the VF parameters were not reported, and the parameters were measured only on some CSC levels. In 2010, Onu et al [1] reported that significant FA differences were observed between ROIs in the VF and DF, LF and DF, and LF and VF. However, their sample size was small (8 cases), and DTI parameters were only measured at C5/6.…”

“…FA and MD were measured at each of these 48 ROIs. The measurements from the both sides of WM funiculi and GM were averaged, and the mean values were used for analysis, as previously reported [1]. The tracing of the ROIs was performed to exclude at least two voxels outside the cord, thereby minimizing the risk of partial volume effects due to the cerebrospinal fluid (CSF).…”

“…The WM can be divided into dorsal, lateral, and ventral funiculi (DF, LF and VF, respectively). Funiculi carry distinct ascending somatosensory and descending motor spinal cord tracts [1]. Previously studies [2][3][4][5][6] revealed that the dorsomedian tract, the Burdach's tract and the medial part of the LF were thin, while the fiber bundles on the outside of the VF and LF were coarse.…”