gms | German Medical Science

Objective: Today visualization of major white matter tracts based on diffusion tensor imaging (DTI) plays an important role during brain tumor surgery, especially glioma surgery. It enables extended tumor resection while minimizing postoperative deficits and thereby lengthens overall survival times and quality of life. Beside other factors like artifacts, quality of white matter tract reconstruction depends on the general quality of the implemented diffusion weighted image data sets considered for diffusion tensor calculation, often described using the signal to noise ratio (SNR). A 32 channel (ch) head coil is capable of improving SNR and thereby delivering higher image quality in comparison to other standard head coils (8ch or 12ch). The effect of using a 32ch head coil instead of a 12ch head coil on white matter tract reconstruction has not been investigated up to now.

Method: MRI data was acquired from 10 healthy volunteers (mean age: 31.00, SD: 7.72, m/f: 5/5) at a 3T-MRI (Trio, Siemens) including a 3D-T2 and two DTI data sets (12ch, 32ch). Imaging parameters were the same for both types of DTI scans. Fiber tracking of the corticospinal tract was performed within the iPlan Software (BrainLab). For this purpose cerebral peduncle and motor cortex were manually segmented within the right hemisphere. Fiber tractography was initiated using the outlined structures with fractional anisotropy (FA) threshold of 0.2 and minimum fiber length of 80mm. After tractography on all DTI data sets, fibers were corrected for parts not belonging to the specified tract by using additional excluded regions. Afterwards, tract specific parameters (mean FA, object volume) were analyzed.

Results: Fiber tractography based on DTI data sets using the 12ch head coil delivered a mean FA value of 0.518 (SD: 0.027) and mean object volumes of 8.321cm³ (SD: 1.933cm³). Using the 32ch head coil reconstructed objects showed a mean FA value of 0.531 (SD: 0.025) and a mean object volume of 8.947cm³ (SD: 2.047cm³). Paired students t-test showed a significant difference between the two groups concerning the mean FA value (p < 0.028), whereas there was no difference found for the object volume.

Conclusions: Fiber tractography and white matter tract characteristics depend on image quality. Increasing image quality and thereby reducing the risks of underestimation of FA values according to lower SNRs is an important factor for reliable white matter tract reconstruction. This can be supported by the use of a 32ch head coil instead of standard head coils.