Fiber Tracking with Distinct Software Tools Results in a Clear Diversity in Anatomical Fiber Tract Portrayal

Abstract

Background: Fiber tract portrayal, based on diffusion tensor imaging (DTI), is becoming more and more important in functional neuronavigation. No standard exists to guarantee anatomically correct fiber tract depiction for neurosurgical purposes. Therefore, showing the anatomically correct extension of fiber tracts beyond the pure connection of functional areas remains an area of important research and investigation. Standards for fiber tracking software applications are elusive. The purpose of this study was to compare the performance of different fiber tracking software tools (FT-tools). We tested the software performance, comparability and anatomical accuracy of the tracking results of several programs.

Material and Methods: A single DTI dataset of a healthy control subject was submitted to four different fiber tracking software applications (two commercial, two freeware), three of them based on Fiber Assignment by Continuous Tracking, one based on the Tensorline Propagation Algorithm. The corticospinal tract (CST) was investigated. The tracking procedure was controlled by the following input variables: single regions of interest (ROIs): brain stem, or internal capsule, or subcortical white matter of the precentral gyrus; background threshold, fractional anisotropy (FA) threshold, maximum fiber angulation and fiber length. Tracking results were compared for 2-D correlated triplanar images (axial, coronal, sagittal) and in 3-D. For all FT-tools, the time used to generate the CST was measured. The inter-rater variability for tracking time and for the tracked CST volumes was recorded for two of the four FT-tools.

Results and Conclusions: Distinct FT-tools performed very differently with respect to the time required to achieve CST portrayal (track generation time varied between 16 and 50 min). None of the software applications was able to display the CST in its full anatomical extent. Especially the lateral precentral areas were not pictured. Surprisingly, the application of the four distinct FT-tools did not lead to comparable tracking results. As very similar or identical tracking algorithms were used, this difference cannot be easily explained. Clearly, neurosurgeons have to be cautious about applying fiber tracking results intraoperatively, especially when dealing with an abnormal or distorted fiber tract anatomy. The authors recommend the use of adjunct strategies such as intraoperative electrophysiology to enhance patient safety and improve anatomical accuracy when using tracking results for surgical procedures.

Zusammenfassung

Hintergrund: Die Abbildung von Faserbahnen mittels der Diffusions-Tensor-Bildgebung (DTI) gewinnt mehr und mehr an Bedeutung für die funktionelle Neuronavigation. Es existieren derzeit weder definierte Standards für die Faserbahndarstellung nach neurochirurgischen Maßstäben, die über die reine Verbindung von Funktionseinheiten des Gehirns hinausgehen und die Faserbahnen in ihrer anatomischen Ausdehnung darstellen, noch für Fiber Tracking Softwareapplikationen. Ziel dieser Studie war der Vergleich verschiedener Tracking Programme (FT-Tools). Als Zielvariablen wurden Anwendbarkeit sowie Vergleichbarkeit und anatomische Genauigkeit der Tracking-Ergebnisse definiert.

Material und Methoden: Der DTI-Datensatz einer gesunden Kontrollperson wurde jeweils mit vier verschiedenen FT-Tools analysiert (zwei kommerzielle, zwei public domain). Drei nutzen den FACT-Algorithmus (Fiber Assignment by Continuous Tracking), eines den Tensorline Propagation Algorithmus. Es wurde der Tractus corticospinalis (CST) untersucht. Das Tracking-Verfahren wurde durch folgende Eingabevariablen kontrolliert: Definition der ROI (Region of interest: Hirnstamm, Capsula interna, subkortikale weiße Substanz des Gyrus präcentralis), Hintergrund-Schwellenwert des ungewichteten Referenzbildes (b0), Fraktionale Anisotropie (FA) -Schwellenwert, maximaler Faserwinkel und Faserlänge. Die Ergebnisse wurden sowohl in 2-D (coronar, axial, sagittal) als auch in 3-D qualitativ verglichen. Für alle Programme wurde die Zeit zur Generierung des CST bestimmt. Weiterhin wurde für zwei der FT-Tools die Interrater-Variabilität (CST-Volumen) analysiert.

Resultate und Schlussfolgerungen: Keines der hier angewandten FT-Tools konnte den CST in seiner vollständigen anatomischen Ausdehnung darstellen. Vor allem wurden die lateralen Anteile der Faserbahn nicht oder nur unvollständig detektiert. Zwischen den FT-Tools fanden sich erhebliche Unterschiede bei der Darstellung von Ausdehnung und Form des CST. Neurochirurgen sollten daher bei der Interpretation und Anwendung solcher Tracking-Ergebnisse im operativen Kontext vorsichtig sein. Mit den getesteten FT-Tools kann daher die Ausdehnung einer Faserbahn möglicherweise falsch eingeschätzt oder unterschätzt werden. Die Autoren empfehlen zusätzliche Strategien, z. B. die Anwendung intraoperativer Elektrophysiologie.

Key words

corticospinal tract - diffusion tensor imaging - neurosurgical fiber tracking - functional neuronavigation - magnetic resonance imaging

Schlüsselwörter

Diffusions-Tensor-Bildgebung - funktionelle Neuronavigation - Magnetresonanztomografie - Tractus corticospinalis

References

• 1 Akai H, Mori H, Aoki S. et al . Diffusion tensor tractography of gliomatosis cerebri: fiber tracking through the tumor.  J Comp Assist Tomogr. 2005;  29 127-129
• 2 Basser PJ, Pierpaoli C. Micro structural and physiological features of tissues elucidated by quantitative diffusion tensor MRI.  J Magn Reson B. 1996;  111 209-219
• 3 Basser PJ, Pajevic S, Pierpaoli C. et al . In vivo fiber tractography using DT-MRI data.  Magn Reson Med. 2000;  44 ((4)) 625-632
• 4 Berman JI, Berger MS, Mukherjee P. et al . Diffusion-tensor imaging-guided tracking of fibers of the pyramidal tract combined with intraoperative cortical stimulation mapping in patients with gliomas.  J Neurosurg. 2004;  101 66-72
• 5 Braun V, Dempf S, Tomczak R. et al . Multimodal cranial neuronavigation: direct integration of functional magnetic resonance imaging and positron emission tomography data: technical note.  Neurosurgery. 2001;  48 1178-1181
• 6 Bürgel U, Amunts K, Hoemke L. et al . White matter fiber tracts of the human brain: Three-dimensional mapping at microscopic resolution, topography and intersubject variability.  Neuroimage. 2006;  29 1092-1105
• 7 Carpenter MB. Core Text of Neuroanatomy. Williams & Wilkins 1991
• 8 Chen X, Weigel D, Gansland O. et al . Diffusion tensor-based fiber tracking and intraoperative neuronavigation for the resection of a brainstem cavernous angioma.  Surg Neurol. 2007;  68 285-291
• 9 Ciccarelli O, Toosy AT, Parker GJ. et al . Diffusion tractography based group mapping of major white-matter pathways in the human brain.  Neuroimage. 2003;  19 1545-1555
• 10 Coenen VA, Krings T, Mayfrank L. et al . 3-D visualization of the pyramidal tract in a neuronavigation system during brain tumor surgery: first experiences and technical note.  Neurosurgery. 2001;  49 86-93
• 11 Coenen VA, Krings T, Weidemann J. et al . Diffusionsgewichtete MRT kombiniert mit navigiertem 3-D-Ultraschall und fMRT zur Entfernung eines Kavernoms der Sehstrahlung.  Zentralbl Neurochir. 2003;  64 133-137
• 12 Coenen VA, Krings T, Weidemann J. et al . Sequential visualization of brain and fiber tract deformation during intracranial surgery with 3-D ultrasound (3-DUS): An approach to evaluate the effect of brain shift.  Neurosurgery. 2005;  56 ((ONS Suppl.1)) ONS133-ONS144
• 13 Coenen VA, Fromm C, Kronenbürger M. et al . Electrophysiological proof of diffusion weighted imaging derived depiction of the deep-seated pyramidal tract in human.  Zentralbl Neurochir. 2006;  67 117-122
• 14 Dejerine J. Anatomie des centres nerveux. Reuff, Paris, France 1901
• 15 Frank LR. Characterization of anisotropy in high angular resolution diffusion-weighted MRI.  Magn Reson Med. 2002;  47 1083-1099
• 16 Fries W, Danek AD, Scheidtmann K. et al . Motor recovery following capsular stroke. Role of descending pathways from multiple motor areas.  Brain. 1993;  108 697-733
• 17 Geschwind N. Disconnexion syndromes in animal and man.  Brain. 1965;  88 237-294
• 18 Geyer S, Ledberg A, Schleicher A. et al . Two different areas within the primary motor cortex of man.  Nature. 1996;  382 805-807
• 19 Holodny AI, Schwartz TH, Ollenschleger M. et al . Tumor involvement of the corticospinal tract: diffusion magnetic resonance tractography with intraoperative correlation.  J Neurosurg. 2001;  95 1082
• 20 Jiang H, Zijl P van, Kim J. et al . DtiStudio: Resource program for diffusion tensor computation and fiber bundle tracking.  Comput Methods Programs Biomed. 2006;  81 106-116
• 21 Jones DK, Horsfield MA, Simmons A. Optimal strategies for measuring diffusion in anisotropic systems by magnetic resonance imaging.  Magn Reson Med. 1999;  42 515-525
• 22 Kinoshita M, Yamada K, Hashimoto N. et al . Fiber-tracking does not accurately estimate size of fiber bundle in pathological condition: initial neurosurgical experience using neuronavigation and subcortical white matter stimulation.  Neuroimage. 2005;  25 424-429
• 23 Kombos T, Suess O, Ciklatekerlio O. et al . Monitoring of intraoperative motor evoked potentials to increase the safety of surgery in and around the motor cortex.  J Neurosurg. 2001;  95 608-614
• 24 Krishnan R, Raabe A, Hattingen E. et al . Functional magnetic resonance imaging-integrated neuronavigation: correlation between lesion-to-motor cortex distance and outcome.  Neurosurgery. 2004;  55 904-915
• 25 LeBihan D, Poupon C, Amadon A. et al . Artifacts and pitfalls in diffusion MRI.  J Magn Reson Imag. 2006;  24 478-488
• 26 Mikuni N, Okada T, Enatsu R. et al . Clinical impact of integrated functional neuronavigation and subcortical electrical stimulation to preserve motor function during resection of brain tumors.  J Neurosurg. 2007;  106 593-598
• 27 Mori S, Crain BJ, Chacko VP. et al . Three-dimensional tracking of axonal projections in the brain by magnetic resonance imaging.  Ann Neurol. 1999;  51 265-269
• 28 Netsch T, Muiswinkel A van. Quantitative evaluation of image-based distortion correction in diffusion tensor imaging.  IEEE Trans Med Imag. 2004;  23 789-798
• 29 Nieuwenhuys A, Voogd J, Huijzen C Van. The Human Central Nervous System: A Synopsis and Atlas. Springer, Berlin, Germany 1988
• 30 Nimsky C, Gansland O, Hastreiter P. et al . Intraoperative diffusion-tensor MR imaging: shifting of white matter tracts during neurosurgical procedures – initial experience.  Radiology. 2005;  234 218-225
• 31 Nimsky C, Grummich P, Sorensen AG. et al . Visualization of the pyramidal tract in glioma surgery by integrating diffusion tensor imaging in functional neuronavigation.  Zentralbl Neurochir. 2005;  66 133-141
• 32 Nimsky C, Ganslandt O, Fahlbusch R. Implementation of fiber tract navigation.  Neurosurgery. 2006;  58 ONS292-ONS304
• 33 Penfield W, Boldrey E. Somatic motor and sensory representation in the cerebral cortex of man as studied by electrical stimulation.  Brain. 1937;  60 389-443
• 34 Rasmussen  Jr  IA, Lindseth F, Rygh OM. et al . Functional neuronavigation combined with intra-operative 3-D ultrasound: initial experiences during surgical resections close to eloquent brain areas and future directions in automatic brain shift compensation of preoperative data.  Acta Neurochir (Wien). 2007;  149 365-378
• 35 Schmahmann JD, Pandya DN, Wang R. et al . Association fibre pathways of the brain: parallel observations from diffusion spectrum imaging and autoradiography.  Brain. 2007;  130 630-653
• 36 Schulder M, Maldjian JA, Liu WC. et al . Functional image-guided surgery of intracranial tumors located in or near the sensorymotor cortex.  J Neurosurg. 1998;  89 412-418
• 37 Suess O, Kombos T, Ciklatekerlio O. et al . Impact of brain shift on intraoperative neurophysiological monitoring with cortical strip electrodes.  Acta Neurochir (Wien). 2002;  144 1279-1289
• 38 Thirion JP. Image matching as a diffusion process: An analogy with Maxwell's demons.  Medical Image Analysis. 1998;  2 243-260
• 39 Toussaint N, Souplet JC, Fillard P. MedINRIA: Medical image navigation and research tool by INRIA. In: Proceedings of the 10th International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI’07), Lecture Notes in Computer Science. Springer Verlag Brisbane, Australia 2007
• 40 Vercauteren T, Pennec X, Perchant A. et al .Non-parametric diffeomorphic image registration with the demons algorithm. In: Proceedings of the 10th International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI’07), Lecture Notes in Computer Science Springer Verlag Brisbane, Australia 2007
• 41 Weinstein DM, Kindlman GL, Lundberg EC. Tensorlines: advection diffusion based propagation through diffusion tensor fields. In: Ebert DS, Gross MH, Hamann B. IEEE Visualization. IEEE Computer Society and ACM, San Francisco 1999: 249-253
• 42 Woods RP, Cherry SR, Mazziotta JC. Rapid automated algorithm for aligning and reslicing PET images.  J Comp Assist Tomogr. 1992;  16 620-633
• 43 Wu JS, Zhou LF, Tang WJ. et al . Clinical evaluation of follow-up outcome of diffusion tensor imaging-based functional neuronavigation: a prospective, controlled study in patients with gliomas involving pyramidal tracts.  Neurosurgery. 2007;  61 935-948
• 44 Xue R, Zijl PCM van, Crain BJ. et al . In vivo three-dimensional reconstruction of rat brain axonal projections by diffusion tensor imaging.  Magn Reson Med. 1999;  42 1123-1127
• 45 Yamada K, Sakai K, Hoogenraad FGC. et al . Multitensor tractography enables better depiction of motor pathways: initial clinical experience using diffusion-weighted MR imaging with standard b-value.  Am J Neuroradiol. 2007;  28 1668-1673

Correspondence