Magnetic Resonance Imaging Diffusion Tensor Tractography: Evaluation of Anatomic Accuracy of Different Fiber Tracking Software Packages

Background Diffusion tensor imaging (DTI)–based tractography has become an integral part of preoperative diagnostic imaging in many neurosurgical centers, and other nonsurgical specialties depend increasingly on DTI tractography as a diagnostic tool. The aim of this study was to analyze the anatomic...

Full description

Saved in:
Bibliographic Details
Published in:World neurosurgery 2014, Vol.81 (1), p.144-150
Main Authors: Feigl, Guenther C, Hiergeist, Wolfgang, Fellner, Claudia, Schebesch, Karl-Michael M, Doenitz, Christian, Finkenzeller, Thomas, Brawanski, Alexander, Schlaier, Juergen
Format: Article
Language:English
Subjects:
Adult
Algorithms
Anatomic accuracy
Anisotropy
Brain mapping
Diffusion Tensor Imaging - methods
DTI
Female
Humans
Image Processing, Computer-Assisted - methods
Male
Middle Aged
Navigation
Nerve Fibers
Neural Pathways
Neurosurgery
Radiology
Reproducibility of Results
Software
Surgery
Young Adult
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Background Diffusion tensor imaging (DTI)–based tractography has become an integral part of preoperative diagnostic imaging in many neurosurgical centers, and other nonsurgical specialties depend increasingly on DTI tractography as a diagnostic tool. The aim of this study was to analyze the anatomic accuracy of visualized white matter fiber pathways using different, readily available DTI tractography software programs. Methods Magnetic resonance imaging scans of the head of 20 healthy volunteers were acquired using a Siemens Symphony TIM 1.5T scanner and a 12-channel head array coil. The standard settings of the scans in this study were 12 diffusion directions and 5-mm slices. The fornices were chosen as an anatomic structure for the comparative fiber tracking. Identical data sets were loaded into nine different fiber tracking packages that used different algorithms. The nine software packages and algorithms used were NeuroQLab (modified tensor deflection [TEND] algorithm), Sörensen DTI task card (modified streamline tracking technique algorithm), Siemens DTI module (modified fourth-order Runge-Kutta algorithm), six different software packages from Trackvis (interpolated streamline algorithm, modified FACT algorithm, second-order Runge-Kutta algorithm, Q-ball [FACT algorithm], tensorline algorithm, Q-ball [second-order Runge-Kutta algorithm]), DTI Query (modified streamline tracking technique algorithm), Medinria (modified TEND algorithm), Brainvoyager (modified TEND algorithm), DTI Studio modified FACT algorithm, and the BrainLab DTI module based on the modified Runge-Kutta algorithm. Three examiners (a neuroradiologist, a magnetic resonance imaging physicist, and a neurosurgeon) served as examiners. They were double-blinded with respect to the test subject and the fiber tracking software used in the presented images. Each examiner evaluated 301 images. The examiners were instructed to evaluate screenshots from the different programs based on two main criteria: (i) anatomic accuracy of the course of the displayed fibers and (ii) number of fibers displayed outside the anatomic boundaries. Results The mean overall grade for anatomic accuracy was 2.2 (range, 1.1–3.6) with a standard deviation (SD) of 0.9. The mean overall grade for incorrectly displayed fibers was 2.5 (range, 1.6–3.5) with a SD of 0.6. The mean grade of the overall program ranking was 2.3 with a SD of 0.6. The overall mean grade of the program ranked number one (NeuroQLab) was 1.7 (range, 1.
ISSN:1878-8750
1878-8769
DOI:10.1016/j.wneu.2013.01.004