Article
Is fiber tract mapping with diffusion tensor MRI accurate enough for neurosurgical procedures?
Ist die Faserbahnabgrenzung im Diffusions-Tensor MRT für neurochirurgische Anwendungen präzise genug?
Search Medline for
Authors
Published: | May 4, 2005 |
---|
Outline
Text
Objective
Fiber tract mapping with DTI is becoming popular for planning neurosurgical approaches and intraoperative neuronavigation. Although an increasing number of studies deal with DTI techniques or neuronal connectivity, there is no proof as to how far DTI corresponds with “real” fiber anatomy. DTI, however, does not directly render the anatomy but reflects a functional condition and it can be susceptible to artifacts, e.g., in pathologically altered brains. Therefoer, we analyzed the accuracy of fiber tract anatomy in DTI by comparing the optic radiation (OR) and corticospinal tract (CT) as delineated in vivo with the tracts as mapped in histological sections of postmortem brains.
Methods
10 human brains were serially sectioned, myelin stained and digitized. In the digitized sections, the tracts were delineated. The image series were 3-D reconstructed and the volume files were linearly and nonlinearly warped to a stereotactic reference space of an in vivo MR-standard brain. The individual fiber tracts were superimposed, and probability maps were generated as a quantitative measure of intersubject variability. DTI scans and high resolution T1 weighted 3-D sequences were obtained in 10 healthy adults. The fiber tracts were identified in each section on the basis of anatomical landmarks, e.g., lateral ventricles (LV), calcarine sulcus, precentral gyrus, internal capsule (IC) and the diffusion images. Finally, the DTI data were warped to the same standard space, and probability maps were calculated.
Results
Comparing the histological map with the DTI probability maps indicates a similar distribution of variability. E.g., variability of the OR is high below the cortical surface and the overlap increases on its course lateral to the LV towards the temporal isthmus. The superimposition of the histological and DTI maps enabled a quantification of the accuracy of the DTI-based mapping: The proportion of the DTI-volumes became more similar to the histological volumes with increasing level of overlap (e.g., the CT in the IC). In defined regions the fiber tracts in DTI „hit“ the histological tract in all probability with an accordanceof up to 90%.
Conclusions
This study shows that DTI- based mapping of the OR and CT is quite precise and fits well with the histology, the present “gold-standard” concerning spatial and grey-value resolution. For clinical routine, interactive fiber tract delineation is time- consuming and only in rare cases does it reveal new aspects which make a change in the surgical proceeding necessary.