Artikel
Comparison of visualized BOLD activation with different fMRI software packages
Suche in Medline nach
Autoren
-
Christian Doenitz
- Klinik und Poliklinik für Neurochirurgie, Universitätsklinikum Regensburg, Germany
-
Claudia Fellner
- Institut für Röntgendiagnostik, Universitätsklinikum Regensburg, Germany
-
Katharina Rosengarth
- Institut für Psychologie, Universitätsklinikum Regensburg, Germany
-
Jürgen Schlaier
- Klinik und Poliklinik für Neurochirurgie, Universitätsklinikum Regensburg, Germany
-
Alexander Brawanski
- Klinik und Poliklinik für Neurochirurgie, Universitätsklinikum Regensburg, Germany
| Veröffentlicht: | 16. September 2010 |
|---|
Gliederung
Text
Objective: Functional magnetic resonance imaging (fMRI) is widely used in preoperative planning nowadays, mostly in glioma surgery close to eloquent areas. We compared the visualized BOLD activation of a simple motor task paradigm in 20 volunteers with three widely used software packages for fMRI postprocessing (Siemens, Brainlab, SPM5). We compared distribution, extent and magnitude in 80 BOLD activations and discuss the differences and their causes.
Methods: Experiments were performed on a 1.5 T system (Magnetom Symphony, Siemens) using a standard 12-channel head array coil. 20 healthy volunteers, 10 men and 10 women, with a mean age of 33 years were included. The paradigm consisted of a simple finger and toe tapping motor task (block design). Four fMRI experiments were acquired per volunteer. TR 3000 ms, TE 50 ms, flip angle 60°, 33 transverse slices with a slice thickness of 3 mm and an in-plane voxel-size of 3 mm x 3 mm. EPI sequences were postprocessed with three software packages (Siemens, Brainlab, SPM5) using similar processing steps, if available. The activated area in the primary motor cortex (M1) was quantified for a fixed t-value (t≤4.0), analyzing distribution, extent and magnitude of BOLD signal.
Results: We detected significant differences of the visualized BOLD signal in all three software packages, even with comparable postprocessing steps and thresholds leading to a variation of distribution, extent and magnitude of the BOLD signal. Critical differences in the software algorhythms concerning coregistering, smoothing and visualization were found and designated changes in these postprocessing steps led to more comparable results.
Conclusions: We could show significant differences of BOLD signal visualization with different, widely used software packages. With fMRI widely used in preoperative planning, understanding of BOLD signal postprocessing, visualization and interpretation is essential. Recent visualization software pretends to display reality. Due to the lack of a Gold Standard for fMRI processing, it is crucial for the neurosurgeon to know the different options of fMRI postprocessing and visualization and their limitations in order to judge for the patients´ benefit.
