Soccer is the only major sport with voluntary unprotected head-to-ball contact. It is crucial to determine if head impact through regular soccer sports training is manifested in brain structure and connectivity, and whether such alterations are due to sustained training per se. Using diffusion tensor imaging, we documented a comprehensive view of soccer players’ brains in a sample of twenty-five right-handed male elite soccer players aged from 18 to 22 years and twenty-five non-athletic controls aged 19 to 24 years. Importantly, none had recalled a history of concussion. We performed a whole-brain tract-based spatial statistical analysis, and a tract-specific probabilistic tractography method to measure differences of white matter properties between groups. Whole-brain integrity analysis showed increased microstructural integrity within the corpus callosum tract in soccer players compared to controls. Further, tract-specific probabilistic tractography revealed that the anterior part of corpus callosum may be the brain structure most relevant to training experience, which may put into perspective prior evidence showing corpus callosum alteration in retired or concussed athletes practicing contact sports. Intriguingly, experience-related alterations showed left hemispheric lateralization of potential early signs of concussion-like effects. This is the first study to-date providing a definitive characterization of soccer-specific experience-related structural alterations. In sum, we concluded that the observed gains and losses may be due to a consequence of engagement in protracted soccer training that incurs prognostic hallmarks associated with minor injury-induced neural inflammation. Soccer is the only major sport with voluntary unprotected head-to-ball contact. It is crucial to determine if head impact through regular soccer sports training is manifested in brain structure and connectivity, and whether such alterations are due to sustained training per se. Using diffusion tensor imaging, we documented a comprehensive view of soccer players’ brains in a sample of twenty-five right-handed male elite soccer players aged from 18 to 22 years and twenty-five non-athletic controls aged 19 to 24 years. Importantly, none had recalled a history of concussion. We performed a whole-brain tract-based spatial statistical analysis, and a tract-specific probabilistic tractography method to measure differences of white matter properties between groups. Whole-brain integrity analysis showed increased microstructural integrity within the corpus callosum tract in soccer players compared to controls. Further, tract-specific probabilistic tractography revealed that the anterior part of corpus callosum may be the brain structure most relevant to training experience, which may put into perspective prior evidence showing corpus callosum alteration in retired or concussed athletes practicing contact sports. Intriguingly, experience-related alterations showed left hemispheric lateralization of potential early signs of concussion-like effects. This is the first study to-date providing a definitive characterization of soccer-specific experience-related structural alterations. In sum, we concluded that the observed gains and losses may be due to a consequence of engagement in protracted soccer training that incurs prognostic hallmarks associated with minor injury-induced neural inflammation. Soccer is the only major sport with voluntary unprotected head-to-ball contact. It is crucial to determine if head impact through regular soccer sports training is manifested in brain structure and connectivity, and whether such alterations are due to sustained training per se. Using diffusion tensor imaging, we documented a comprehensive view of soccer players’ brains in a sample of twenty-five right-handed male elite soccer players aged from 18 to 22 years and twenty-five non-athletic controls aged 19 to 24 years. Importantly, none had recalled a history of concussion. We performed a whole-brain tract-based spatial statistical analysis, and a tract-specific probabilistic tractography method to measure differences of white matter properties between groups. Whole-brain integrity analysis showed increased microstructural integrity within the corpus callosum tract in soccer players compared to controls. Further, tract-specific probabilistic tractography revealed that the anterior part of corpus callosum may be the brain structure most relevant to training experience, which may put into perspective prior evidence showing corpus callosum alteration in retired or concussed athletes practicing contact sports. Intriguingly, experience-related alterations showed left hemispheric lateralization of potential early signs of concussion-like effects. This is the first study to-date providing a definitive characterization of soccer-specific experience-related structural alterations. In sum, we concluded that the observed gains and losses may be due to a consequence of engagement in protracted soccer training that incurs prognostic hallmarks associated with minor injury-induced neural inflammation.

MRI images were collected on a GE MR750 3T scanner (GE Healthcare, Waukesha, WI, USA) in the Mind Research and Imaging Center of NCKU with a 32-channel brain array coil. All participants were scanned using the same MR scanner. Diffusion-weighted spin-echo echo-planar imaging sequence images were obtained with a measured spatial resolution of 2.5 x 2.5 x 2.5 mm (acquisition matrix 100 x 100 pixels, 50 slices) and a reconstructed resolution of 1.56 x 1.56 x 2.0 mm (reconstructed matrix 100 x 100 pixels, 50 slices). The sequence parameters were repetition time (TR) = 5500ms, echo time (TE) = 62~64 ms, 50 non-linear diffusion directions with b = 1000s/mm2, field of view (FOV) = 250×250 mm2, number of excitations (NEX) = 3, and slice thickness = 2.5 mm. Reverse DTI was also acquired for top-up correction in the DTI preprocessing. The acquisition parameters for the reverse DTI were identical to the DTI except for that only six directions were obtained due to the time constraints. The total acquisition time was 15 minutes 24 seconds.
Three-dimensional high-resolution brain structural images were also acquired using a T1-weighted fast spoiled gradient-echo dual-echo (FSPGR) sequence for each participant to allow for spatial normalization and visualization. The sequence parameters were TR = 2900 ms; TE = 7.6 ms; matrix size 224 x 224; flip angle = 12°; 1 mm slice thickness; FOV = 22.4/1 (cm/Phase); receiver bandwidth (BW) = ±31.25 kHz, and 170 sagittal slices covering the whole brain were collected. The total acquisition time was 3 minutes 38 seconds. For each participant, all the diffusion and anatomical images were acquired in the same session.

All data processing was used to process imaging data by the FMRIB Software Library 5.0 (FSL, https://fsl.fmrib.ox.ac.uk/fsl) (Jenkinson, Beckmann, Behrens, Woolrich, & Smith, 2012) on a Linux platform. NIFTI files are in combined (.nii.gz) format and as a zip file containing separate .img and .hdr files. Detail analysis procedures and steps can be referred to Yao et al. (2020) The brains of elite soccer players are subject to experience-dependent alterations in white matter connectivity. Cortex.

====DTI analysis logsheet===

https://fsl.fmrib.ox.ac.uk/fslcourse/lectures/fdt1.pdf

for the advanced tractography analysis on white matter tracts (including mask images) can be referred to: https://fsl.fmrib.ox.ac.uk/fsl/fslwiki/AutoPtx

1. listed and moved all the particiapnts to the target folder (equal sample size)

naming format=Wxxx(athlete); Yxxx(control); for detail: refer to a .ods file (named: Cortex-dMRIsoccer-25_demographic-INFO)

2.convert the diffusion data from dicom to nifit
run"./dti4dcm $filename of subject type $folder name of original DTI data" (see the script before you execute)

3.preform pre-processing such as distortion correction...
Check script "proc4dti" to perform this for TBSS analysis

4.then try the same correction for future TBSS analysis in different DTI scalars
check script "mv-scalar2dir4tbss"

5.now you have all kinds of diffusion scalars (ready to compare)
check "tbss_run" <- this script did not wriiten for execute mainly listed all the steps you have to go through
make sure you have all files in the target folder

6.after above steps, you can try tractography analysis
for the whole brain tracts: running "run_tractogrphy" <-this step contain same procedures as below but focus on tracts
then run "autoPtx_2_launchTractography"
then run "autoPtx_prepareForDisplay"

IF you prefer to do seed based tracking...

Steps:
you have to modelling all the fiber for tracking first
note: this step is extremly time consuming...
check "run_bedpostx"

then execute "run_X2probtrack" <-still working on it...
run_launchAutoPtx2 <- this is for 27 tracts tracking

VBM analysis -->http://web.mit.edu/fsl_v5.0.8/fsl/doc/wiki/FSLVBM(2f)UserGuide.html
https://fsl.fmrib.ox.ac.uk/fslcourse/lectures/practicals/seg_struc
1. VBM-->mkdir vbm--->move all T1 BET and non-BET to vbm folder by using vbm_move_T1
2. making a "template_list" contain all subject name of T1 with equal group
3. check all "slicesdir `imglob *`"
4. creat glm design matrix
5. type fslvbm_1_bet -b
6. fslvbm_2_template
7. fslvbm_2_template
8. randomise -i GM_mod_merg_s3 -m GM_mask -o fslvbm -d design.mat -t design.con -T -n 5000
9.run vbm_result plot
10. extract intensity of grey matter (VBM stats folder)
-->fsl5.0-fslmeants -i stats/GM_mod_merg_s3.nii.gz -m stats/GM_mask.nii.gz
-->fsl5.0-fslstats stats/GM_mod_merg_s3.nii.gz -k stats/GM_mask.nii.gz -V | cut -d" " -f2
-->to know each subject value -> total volume / n * intensity value for every subject at the fslmeants output
-->fsl5.0-cluster -i vbm-grp-diff_tfce_corrp_tstat1.nii.gz -t 0.95 -o vbm-grp-diff-tstats1 #creat mak for significant cluster from ramdomise results
-->if you want to see certain cluster only use as below: (https://fsl.fmrib.ox.ac.uk/fsl/fslwiki/Cluster)
fsl5.0-fslmaths -dt int yr-ath-tstats1.nii.gz -thr 11 -utr 11 -bin yr-ath-tstats1_cluster_mask11
-->extract value Z value or Beta or PE value by fslmeants
fslmeants -i stats/GM_mod_merg_s3.nii.gz -m vbm-grp-diff-tstats1.nii.gz