Design
A retrospective study in radiology.

Time and setting
Experiments were performed in the Nuclear Magnetic Resonance Laboratory, Department of Neurosurgery, Chinese PLA General Hospital, China from April 2010 to April 2012.

Subjects
We recruited 185 patients undergoing surgery for glioma (diagnosis confirmed by postoperative pathological examination) involving the left frontal lobe at the Chinese PLA General Hospital. Among the 185 patients, 10 patients fulfilled the following inclusion criteria: (1) patients underwent diffusion tensor imaging scanning in our hospital preoperatively; (2) the tumor location was confined to the left frontal region above the level of the superior insular fissure and anterior to the posterior edge of the corpus callosum (as shown in Figure 5); (3) the functional MRI and Edinburgh handedness index indicated left language dominance^[34]; (4) preoperative language function was normal (aphasia quotients > 93.8)^[35]; (5) the tumor diameter was more than 3 cm; (6) pathological diagnosis revealed WHO II grade glioma. Exclusion criteria: patients with disease that may affect white matter, such as multiple sclerosis, dementia and schizophrenia were excluded.

Of the 10 included patients, seven were male and three were female, with ages ranging from 35 to 54 years. A total of 10 normal right-handed controls (from the healthy check-up population) were also tested (seven males and three females, ages ranging from 38 to 49 years). Handedness was determined using the Edinburgh handedness inventory. Informed consent was provided by participants and family members, and the study was conducted in accordance with the Declaration of Helsinki.

Methods
Image acquisition
Diffusion tensor imaging was performed with a 1.5 T scanner (Siemens Espree, Erlangen, Germany). We used a single-shot spin-echo diffusion weighted echo-planar imaging sequence^[36]. The sequence parameters were as follows: echo time 147 ms, repetition time 9 400 ms, matrix 128 × 128, field of view 232 mm × 232 mm, slice thickness 2.7 mm. 12 directions of diffusion-weighted imaging, b₁ (high b value) 1 000 s/mm², b0 (low b value) 0 s/mm², voxel size 1.8 mm × 1.8 mm × 2.7 mm. Forty-five layers were continuously acquired four times to improve the signal-to-noise ratio. Co-registered magnetization-prepared rapid gradient echo images were recorded for anatomical guidance.

Image processing
Diffusion tensor imaging datasets were transferred to a PC running Windows. The DICOM files were converted to the 4D NifTI image format for analysis using dcm2nii software (University of South Carolina, Columbia, NJ, USA, www.mricro.com). We used diffusion tensor imaging studio (Hangyi Jiang and Susumu Mori, Johns Hopkins University and Kennedy Krieger institute, http://godzilla. kennedykrieger.org or http://lbam.med. jhmi.edu) analysis software to process the data. Images were first realigned using the Automatic Image Registration command to remove any potential small bulk motions that occurred during the scans. All diffusion-weighted images were then visually inspected by the researchers for apparent artifacts due to subject motion and instrument malfunction. The six elements of the diffusion tensor were calculated for each pixel using multivariate linear fitting^[29]. After tensor diagonalization, three eigenvalues and eigenvectors were obtained and fractional anisotropy maps were calculated. The eigenvector associated with the largest eigenvalue was used as an indicator for fiber orientation. In the diffusion tensor imaging color maps, red, green, and blue colors were assigned to right-left, anterior-posterior, and superior-inferior orientations, respectively.

Tractography
The arcuate fasciculus, inferior longitudinal fasciculus and inferior fronto-occipital fasciculus were reconstructed. Fiber assignment was performed with a continuous tracking algorithm^[37] in this study with a fractional anisotropy threshold of 0.2 and an inner product threshold of 0.70. Fiber tracking was performed using diffusion tensor imaging studio (Johns Hopkins University and Kennedy Krieger institute, Baltimore, MD, USA). A multi-region of interest approach was used to reconstruct tracts of interest, exploiting existing anatomical knowledge of tract trajectories. Tracking was performed from all pixels inside the brain (brute-force approach) and results penetrating the manually defined regions of interest were assigned to the specific tracts associated with the regions of interest. We defined the regions of interest according to previously reported protocols^[27]. After tracts were reconstructed, mean fractional anisotropy was calculated over the entire tract. The lateralization index was used to express relative change of fractional anisotropy values between hemispheres. The lateralization index was calculated from the following equations^[38]:
Lateralization index of fractional anisotropy = 100 × (left fractional anisotropy – right fractional anisotropy)/(left fractional anisotropy + right fractional anisotropy)

Statistical analysis
All data were analyzed using SPSS 17.0 software (SPSS, Chicago, IL, USA). Fractional anisotropy and lateralization index data were presented as mean ± SD. The independent-samples t-test and chi-square test were used for comparisons between groups. Paired-samples t-tests were used for comparison between hemispheres. A P-value less than 0.05 (two-sided) was considered statistically significant.

Acknowledgments: We would like to thank Xinguang Yu, Jun Zhang, Zhenghui Sun, Jinli Jiang, Xiaodong Ma, Bo Bu and Ruyuan Zhu from Department of Neurosurgery, Chinese PLA General Hospital for their collaborative support.
Funding: This work was supported by the National Natural Science Foundation of China, No. 31040039; the Natural Science Foundation of Beijing, No. 7102145; and the Military Clinical High-Tech Foundation, No. 2010gxjso94.
Author contributions: Gang Zheng was responsible for data integration and analysis, writing the manuscript, and performing statistical analysis. Bainan Xu and Xiaolei Chen participated in the research concept and design, critiqued the manuscript and headed the fund. Jiashu Zhang, Xueming Lv, Jinjiang Li, Fangye Li, Shen Hu, Ting Zhang, and Ye Li provided technical or material support, as well as research guidance. All authors approved the final version of the paper.
Conflicts of interest: None declared.
Ethical approval: The pilot project was approved by the Medical Ethics Committee of Chinese PLA General Hospital, China.
Author statements: The manuscript is original, has not been submitted to or is not under consideration by another publication, has not been previously published in any language or any form, including electronic, and contains no disclosure of confidential information or authorship/patent application/funding source disputations.

null

null