Introduction

Long-term intensive training has been shown to increase the unparalleled athletic ability and might induce brain plasticity in athletes.1, 2 A promising technique called connectome has been introduced as a method to evaluate brain plasticity.3The connectome is a network representation of whole-brain connectivity, which can be mapped to reveal brain circuit-based alterations.4Even though connectome analysis using deterministic tractography has been used to evaluate the brain structure of gymnasts,3deterministic tracking was developed based on unimodal diffusion tensor imaging making it challenging to estimate neural fiber connections in a voxel in which there are crossing or kissing fibers.5In contrast, probabilistic tractography algorithm was proposed to overcome this limitation by estimating multiple fiber directions. The uses of probabilistic tractography with a new method called multi-shell multi-tissue constrained spherical deconvolution (MSMT-CSD)6based on multi-shell diffusion MRIhas been shown to increase the precision of tractography compare to single-shell, single tissue-CSD. In this study, we aimed to clarify how the brain network structure of gymnasts is different from that of ordinary people using connectome analysis with MSMT-CSD.

Materials and Methods

Ten world-class gymnasts (all men; age, mean ± SD 19.9 ± 1.3 years) and ten controls (all men, age, 20.6 ± 1.7 years) were recruited for this study. Each gymnast had trained for a long-term (13.6 ± 2.2 years) and won at least one prize in the world competition or national tournament.Neuroimaging data were obtained on a 3-T system (MAGNETOM Prisma; Siemens Healthcare) with a 64-channel head coil. Diffusion-weighted images were acquired at b-values of 1000 and 2000 s/mm2along 32 uniformly distributed directions with spin-echo echo-planar imaging. Three-dimensional high-resolution sequence MPRAGE (Magnetization Prepared Rapid Acquisition with Gradient-echo) was also obtained. Further, we acquired 84 nodes according to Desikan-Killiany cortical atlas segmentation (FreeSurfer). Whole-brain tractograms were generated using probabilistic MSMT-CSD tracking method with the MRtrix software. The number of streamlines between these nodes were counted as a weighted of connectome. This resulted in an 84 × 84 interregional connectivity matrix. We adopted the Network-based statistics (NBS) for detecting some alterations of WM connectivity by long-term intensive motor training. Topological measures based on graph theory were also calculated using Brain Connectivity Toolbox. In addition, we performed a Pearson’s correlation analysis between the nodal strength and Difficulty-score (D-score) which represents the difficulty of gymnastic apparatus. Finally, we defined the twelve regions with the highest nodal strength value as the hub regions.9

Results

The NBS identified significantly increased subnetworks connectivity comprising 13 edges and 14 nodes in gymnasts group relative to control groups (P< 0.01, Fig. 1). In global network measures, we detected significantly decreased characteristic path length and increased strength in gymnasts compared to control groups (Table 1). In the gymnast group, we detected significantly increased nodal strength in the rostral middle frontal (RMF) that was localized in associative regions and in the bilateral temporal poles that were localized in the limbic region (Table 2). We identified hubs as twelve regions with the highest strength values in the anatomical networks of each group (Figure 2). Among them, right RMF was only detected in the gymnasts. A Pearson’s correlation analysis detected significant positive correlations between nodal strength and D-score (Table 3).

Discussion and Conclusion

We compared anatomical brain networks between world-class gymnasts and controls using the probabilistic MSMT-CSD method. We detected increased structural connectivities in the motor, associative, visual and limbic areas in gymnasts compared to control groups. These specific brain network changes might reflect the brain plasticity induced by long-term intensive training. In the graph theory analysis, the gymnast group displayed increased nodal strength in the bilateral temporal pole related to visuospatial perception and in the RMF related to motor planning that are the functions required for gymnastics. We also detected changes in RMF in the hub region in gymnast group. Further, we detected significant positive correlation between D-score and nodal strength of auditory, visual, motor, limbic and associative area. As mentioned above, the functions of these areas play an essential role in gymnastics. In addition, a significant positive correlation between D-score and nodal strength was observed predominantly in the auditory area. As has been demonstrated previously, spatial cues provided by auditory information might be integrated to obtain a better postural control. In conclusion, our findings can be useful for a better understanding of neural changes related to gymnastic skills.

References

1. Nakata H, Yoshie M, Miura A, Kudo K. Characteristics of the athletes' brain: evidence from neurophysiology and neuroimaging. Brain research reviews. 2010;62(2):197-211.2. Dayan E, Cohen LG. Neuroplasticity subserving motor skill learning. Neuron. 2011;72(3):443-54.3. Wang B, Fan Y, Lu M, Li S, Song Z, Peng X, et al. Brain anatomical networks in world class gymnasts: a DTI tractography study. NeuroImage. 2013;65:476-87.4. Fornito A, Zalesky A, Breakspear M. The connectomics of brain disorders. Nature reviews Neuroscience. 2015;16(3):159-72.5.Basser P.J., Pajevic S., Pierpaoli C., Duda J., Aldroubi A. In vivo fiber tractography using DT-MRI data. Magn. Reson. Med. 2000;44(4):625–632.6. Jeurissen B, Tournier JD, Dhollander T, Connelly A, Sijbers J. Multi-tissue constrained spherical deconvolution for improved analysis of multi-shell diffusion MRI data. NeuroImage. 2014;103:411-26.7. Blaizot X, Mansilla F, Insausti AM, Constans JM, Salinas-Alaman A, Pro-Sistiaga P, et al. The human parahippocampal region: I. Temporal pole cytoarchitectonic and MRI correlation. Cerebral cortex. 2010;20(9):2198-212.8. Hale JB, Fiorello CA. School neuropsychology: A practitioner's handbook: Guilford Publications; 2017.9.van den Heuvel, M.P., Sporns, O., 2013. Network hubs in the human brain. Trends Cogn Sci 17, 683-696.