Fast and Accurate Semi-Automated Segmentation Method of Spinal Cord MR Images at 3T Applied to the Construction of a Cervical Spinal Cord Template

Abstract

ObjectiveTo design a fast and accurate semi-automated segmentation method for spinal cord 3T MR images and to construct a template of the cervical spinal cord.Materials and MethodsA semi-automated double threshold-based method (DTbM) was proposed enabling both cross-sectional and volumetric measures from 3D T2-weighted turbo spin echo MR scans of the spinal cord at 3T. Eighty-two healthy subjects, 10 patients with amyotrophic lateral sclerosis, 10 with spinal muscular atrophy and 10 with spinal cord injuries were studied. DTbM was compared with active surface method (ASM), threshold-based method (TbM) and manual outlining (ground truth). Accuracy of segmentations was scored visually by a radiologist in cervical and thoracic cord regions. Accuracy was also quantified at the cervical and thoracic levels as well as at C2 vertebral level. To construct a cervical template from healthy subjects’ images (n=59), a standardization pipeline was designed leading to well-centered straight spinal cord images and accurate probability tissue map.ResultsVisual scoring showed better performance for DTbM than for ASM. Mean Dice similarity coefficient (DSC) was 95.71% for DTbM and 90.78% for ASM at the cervical level and 94.27% for DTbM and 89.93% for ASM at the thoracic level. Finally, at C2 vertebral level, mean DSC was 97.98% for DTbM compared with 98.02% for TbM and 96.76% for ASM. DTbM showed similar accuracy compared with TbM, but with the advantage of limited manual interaction.ConclusionA semi-automated segmentation method with limited manual intervention was introduced and validated on 3T images, enabling the construction of a cervical spinal cord template.

Analysis of diffusion tensor measurements of the human cervical spinal cord based on semiautomatic segmentation of the white and gray matter

“…Of the aforementioned studies, the CLASS method presented in this article achieves the most accurate results in segmenting the entire spinal cord from a T 2 image. The more recent studies and studies dealing with WM and GM segmentation 5,7,9,[14][15][16][17][18][19]38 use mostly MR with 3T induction. T 2 -weighted images from such MR devices have better signal-to-noise ratios, and thus the contrast between GM and WM is sharper and the segmentation is more accurate.…”

“…There exist many methods for segmenting brain images, but there are far fewer methods for spinal cord segmentation . Most of these methods segment the entire spinal cord (ESC) and CSF at various levels of automation (semiautomatic or fully automatic) and from various image modalities (T 1 , T 2 , DTI, and others) . Only a few methods are able to segment not just the ESC but also WM and GM …”

“…4 Most of these methods segment the entire spinal cord (ESC) and CSF at various levels of automation (semiautomatic or fully automatic) and from various image modalities (T 1 , T 2 , DTI, and others). [5][6][7][8][9][10][11] Only a few methods are able to segment not just the ESC but also WM and GM. 7,9,[12][13][14][15][16][17][18][19] The main objective of this study was to evaluate the possibilities for using a semiautomatic segmentation method (CLASS) based on a semisupervised machine-learning technique implemented in ITK-SNAP [20][21][22][23][24] for segmentation of MRI data of the cervical spinal cord and to utilize this technique for analyzing diffusion parameters of different spinal cord tissues.…”

Deciphering spreading mechanisms in amyotrophic lateral sclerosis

“…A study using multiparametric MRI, combined with evoked sensory potentials, established that the spinal cord lesions extend beyond the motor system to the sensory fibers [24]. The development of spinal cord templates allowing voxel-based morphometric data [25,26] or of spinal cord functional MRI [27] will offer new tools to investigate in vivo the kinetics of ALS lesion propagation within the spinal cord.…”

Topologically preserving straightening of spinal cord MRI

“…. Generic MRI templates and atlases of the spinal cord have been developed as well, facilitating multicenter studies of large groups of subjects and patients and morphological analysis such as voxel‐based morphometry . Because the majority of spinal cord templates show a straight spinal cord, a key element for template‐based analyses is the need to straighten images during preprocessing.…”