Table 4.6 and indicated no significant between-hemispheres differences for all the proportional thresh-olds: 20% (p= 0.986), 25% (p= 0.327), 30% (p= 0.290), 35% (p= 0.507) and 40% (p= 0.914).
Despite not significantly, the mean small-worldness was lower in the lesioned hemisphere, for all the densities (Table 4.1). These results suggested a greater balance of the local processing and global integration of the non lesioned network. The functional reorganization of the lesioned network was not more constructive when analysing its ability to balance the global integration and local processing and comparing it to the non lesioned network. Therefore, the lesion-induced disturbances seemed to (greater) disturb the hemisphere where the lesion was located. However, the functional reorganization of the lesioned network at density of 40% proposed a plasticity occurring in the lesioned network regarding the global and local connectednesses. Moreover, an improvement of the lesioned network in balancing the global integration and the local specialization was inspected.
efficiency was higher. Quantitatively, both lesioned and non lesioned global efficiency results were concordant to the results from the previous studies; however, for the highest densities the higher results were more similar to the findings of Park et al. [38].
The distribution of the non lesioned global efficiency across the different thresholds was concordant to the distributions of both lesioned and healthy groups excluding the possibility for conclusions about the lesion-induced disturbances reaching the non lesioned hemisphere and affecting its global efficiency.
This inference was expected due to the association of global efficiency to the characteristic path length.
The studies conducted by Huang et al. [39] and Park et al. [38] obtained an increase of the clustering coefficient as more connections were included in both lesioned and healthy networks. The opposite was inspected for the clustering coefficient results acquired by Wang et al. [40] (decreased across densities).
They disclosed that the lesioned networks were less locally interconnected being the local information processing capacity weaker. The clustering coefficient distribution of lesioned and non lesioned hemi-spheres, obtained in this study, was concordant to the findings of Huang et al. [39] and Park et al. [38].
Huang et al. [39] found that the patients group exhibited lower clustering coefficient values than HCs, for all densities, while the clustering coefficient was preserved in the study conducted by Park et al. [38].
Despite that, a slightly higher clustering coefficient of the lesioned group was inspected, for the lower densities. The lesioned and non lesioned clustering coefficient from this study were in concordance to the results obtained by Huang et al. [39], until density of 35%. From density of 35%, the lesioned network seemed to exhibit a stronger local processing capacity. Quantitatively, the non lesioned results were accordant to the clustering coefficient of the healthy group while for the lesioned hemisphere, the clustering coefficient was higher than the results of the lesioned group. The results of this study were also concordant to the results from the study conducted by Park et al. [38]; however their increased was lower.
The distribution of the non lesioned clustering coefficient across densities was in accordance to the distributions of both lesioned and healthy groups for the results obtained by Huang et al. [39] and Park et al. [38] precluding conclusions on whether the lesion-induced disturbances reached the non lesioned hemisphere affecting the clustering coefficient as it occurred for the lesioned hemisphere. The distri-bution of the clustering coefficient for the non lesioned hemisphere was not concordant to either the distribution of clustering coefficient for lesioned or healthy groups from the study conducted by Wang et al. [40]. Thereby, no further information about the disturbances caused by the lesion reaching the non lesioned hemisphere and affecting the clustering coefficient findings could be inferred.
From the nodal results obtained by Huang et al. [39] alterations of clustering coefficient in pre-operative LGG at density of 20% included the OLF (olfactory cortex), HES (Heschl’s gyrus), and CAL (calcarine fissure and surrounding cortex) regions, respectively located in subcortical, temporal and oc-cipital brain areas. The nodes where significant differences in the clustering coefficient between lesioned and non lesioned hemispheres were found at density of 20% involved the ANG, HIP (hippocampus) and THA regions which are respectively located in parietal-(pre)motor, medial temporal and subcortical ar-eas. The difference in the nodal results may be related to the differences in the lesions’ location in the brain. Their lesions were merely located in the frontal lobe while the lesions of the patients analysed in this study involved other regions and pathologies.
The local efficiency increased for both lesioned and healthy groups of the study conducted by Huang et al. [39] and Park et al. [38], as more connections were included in the networks. These results were consistent to the corresponding clustering coefficient results revealing a denser local connectedness of the brain functional networks as the density increased. Wang et al. [40] obtained a decrease in the local efficiency across densities, for both lesioned and healthy groups. Therefore, the local functional
reorganization of networks was not constructive leading to a sparse local connectedness and hence a decrease in the efficiency of local information transfer. The lesioned and non lesioned results from this study were concordant to the local efficiency increase across thresholds inspected by Huang et al. [39]
and Park et al. [38]. Furthermore, they were all quantitatively concordant.
Huang et al. [39] found that the local efficiency was lower for the lesioned groups which was in accordance to the clustering coefficient results. Moreover, Park et al. [38] inspected that the lesioned local efficiency was preserved for all densities, except for 16%, 18% and 24% where the lesioned values were significantly higher than that of HCs. The results from this study were concordant to the results obtained by Huang et al. [39] until density of 40%. At this highest density, the local efficiency of the lesioned hemisphere was concordant to the results for the lesioned group from the study conducted by Park et al. [38] (higher local efficiency).
The distribution of the non lesioned local efficiency across the proportional thresholds was accordant to the distributions of both lesioned and healthy groups for the results obtained by Huang et al. [39]
and Park et al. [38] which precluded to conclude about the lesion-induced disturbances had reached the non lesioned hemisphere affecting its local efficiency. The distribution of the local efficiency for the non lesioned hemisphere was not concordant to either the distribution of local efficiency for lesioned or healthy groups from the study conducted by Wang et al. [40]. Thus, further conclusions relating the influence of the lesion in the local efficiency of the non lesioned hemisphere could not be inferred.
In the studies conducted by Huang et al. [39], Park et al. [38] and Wang et al. [40], as more connec-tions were included in both healthy and lesioned networks, the small-worldness decreased. This showed that the networks tended to be more random. However, an increase of the local processing of the networks was suggested to maintain the small-world property. It should be noted that the results from the study conducted by Wang et al. [40] indicated a decrease of the small-worldness as more connections were in-cluded but the small-worldness did not tended to be preserved. The small-worldness distribution of both lesioned and non lesioned hemispheres, obtained in this study, was concordant to the previous findings of the small-worldness preservation. However, Huang et al. [39] found a higher small-property of LGG groups which showed the changes in the overall organization of the lesioned networks compared to HCs.
Quantitatively, both lesioned and non lesioned small-worldness results from this study were concordant to the small-worldness of HCs. Park et al. [38] also demonstrated a conserved small-world topology of the lesioned group comparing to HCs. Even so, the small-worldness of both lesioned and healthy groups was higher than the results obtained in this study, for lesioned and non lesioned hemispheres. Wang et al. [40] obtained a decreased small-world property in the lesioned groups. Quantitatively, the results of lesioned and non lesioned hemispheres were in accordance to the results of healthy and lesioned groups.
The distribution of the non lesioned small-worldness across densities was concordant to the distri-butions of both lesioned and healthy groups. Then, it was not possible to disclose whether the lesion-induced perturbations disturbed the non lesioned small-worldness as it seemed to affect it in the lesioned hemisphere.
In conclusion, the results from this study seemed to be consistent and concordant to the previous studies, being the major differences involving the higher global and local efficiencies of the lesioned network as well as the increase of the lesioned small-worldness at density of 40%. Explanations for these differences may include differences in the investigation’s protocol and pathologies of the patients.
Conclusion
Although the lesion hemispheric location was unilateral, both hemispheres could be affected by the lesion, through short- and long-distance within the lesioned hemisphere. Furthermore, between-hemispheres functional interactions reaching the non lesioned hemisphere might had been associated to lesion-induced disturbances in this hemisphere.
In this perspective, the lesioned and non lesioned hemispheres were analysed in terms of their integra-tion and segregaintegra-tion topological properties to conclude about the lesion’s influence in both hemispheres.
Furthermore, the effect of thresholding in those network properties was also evaluated.
For the integration analysis, it was demonstrated that as more connections were included in the networks, the path lengths between regions tended to be shorter. This suggested an increase of the global integration indicating that the communication between distributed regions seem to be more facilitated.
The global efficiency results corroborated these findings (increased as more connections were included).
Moreover, almost all nodes seemed to significantly contributed for this increase due to the significant differences that were inspected between all pairs of densities.
The functional reorganization of the networks revealed to be more efficient in the non lesioned hemi-sphere at all densities, except at the highest density of 40%, which would imply a greater effective integrity and a more rapid information propagation between and across different regions. This result was concordant to the shorter path lengths of the non lesioned network at all densities, except at the lowest density of 20%. At the density of 40%, the lesioned network seemed to be more globally integrated.
The analysis of the segregation for both lesioned and non lesioned networks involved the cluster-ing coefficient and local efficiency findcluster-ings that respectively measured the level of local combination (connectedness) and propagation (efficiency) of the information. The between-threshold results of these graph measures showed that as more connections were considered in the networks, the level of local connectedness increased inducing a greater local information processing. The lesioned clustering co-efficient values were lower until density of 30% (inclusive) suggesting a lower potential for the local information transfer. These results were concordant to the lower local efficiency for the lesioned net-work. From density of 35%, the clustering coefficient of the lesioned network was higher; however, the local efficiency results indicated a lower efficiency in the local information transfer. At density of 40%, the lesioned local efficiency increased enough to exceed the non lesioned results which proposed a higher local specialization in the lesioned network.
These results showed that the lesioned network may had compensated the lesion-induced perturba-tions within this hemisphere through the establishment of compensatory paths and local mechanisms.
Even so, this conclusion could merely be deduced with the functional reorganization occurred when the lesioned network was thresholded at density of 40%.
Further, this study demonstrated that both lesioned and non lesioned networks revealed a small-world topological organization which enlighten the networks’ ability for specialized processing to occur within densely interconnected modules and to combine information for distributed brain regions. However, as more connections were included in the networks, their ability to balance the global integration and the local processing decreased. At density of 40%, the lesioned network reflected a constructive reorganiza-tion for the local processing and global integrareorganiza-tion. Despite the compensatory mechanisms, the lesioned network exhibited a less optimal small-world organization which suggested a lower capacity in balancing the local information processing in respect to the global integration.
Inconsistencies were found at densities of 20% and 40%, for the integration results. The segregation results were only inconsistent at density of 35%. As explained in Part 2.3.5 of Background, despite the proportional thresholds seeming to improve the network stability, the instability of network measures within reasonable proportional thresholds was expected. Furthermore, it should be noted that the outliers of the data were not excluded from these analyses due to the small number of subjects. Thus, these results may also be caused by the small number of subjects who participated in this study that could had exaggerated or minimized the differences between lesioned and non lesioned networks and between thresholds. The inconsistent results obtained through the comparison of the mean values were expected due to their higher susceptibility for the outliers. Besides, the fact that the nodal results corresponded to regions that were scattered over different brain-hemispheric locations may had been a consequence from the variability of the pathologies within the group of patients: brain tumors, with different locations in the hemisphere, and a case of a cavernous malformation.
In summary, the results seemed to be sensitive to the choice of threshold. Although instabilities had been inspected in the results of this study, the selection of reasonable ranges of proportional thresholds is proposed. Therefore, it would allow for more stable networks highlighting the robustness of the results.
The lesioned hemisphere showed a lower integration and segregation than the non lesioned hemi-sphere, when excluding the analysis at the higher thresholds. This reflected that indeed the lesion-induced alterations affected the functional connectivity within the hemisphere where the lesion is located. At den-sity of 40%, the functional reorganization of the lesioned network was constructive suggesting that the lesion-induced perturbation within this hemisphere were compensated through compensatory paths and local mechanisms. The functional between-hemispheres interactions could also had disturb the non le-sioned hemisphere. However, from the results of this study no clear conclusion could be inferred. In addition, the comparison to previous investigations that recruited patients with lesions and healthy sub-jects precluded in inferring the lesion’s impact in the non lesioned hemisphere.
In future studies it might be interesting to compute more additional graph measures such as the nodal degree to conclude, more accurately, about the nodal effect of including more connections (thresholding) and the integration characteristics of the networks. For the segregation conclusions, the nodal local efficiency can be also computed upgrading and consolidating the findings of the local efficiency.
The concordance between the results of this study and the findings from previous studies that went further and performed a neuro-cognitive analysis opens up for a future study of the functional cognitive performance integrating data and the results of this study. This will allow the evaluation of whether and which brain lesioned and non lesioned areas are affected by the lesion and the level of this disturbance in terms of cognitive functioning. Furthermore, a further comparison involving the non lesioned hemi-sphere of patients and healthy subjects might be useful to disclose the impact of the lesion in the non lesioned network. Additionally, in the current study, it was investigated and compared the lesioned and non lesioned graph metrics prior to surgery. Future studies may also focus on the establishment of an association between the topological parameters of the networks after the lesion resection and at long
term, and further include a link to the cognitive performance.
[1] J Ferlay et al. “Cancer incidence and mortality patterns in Europe: Estimates for 40 countries and 25 major cancers in 2018”. In:European Journal of Cancer103 (2018), pp. 356–387.
[2] Raul Mattassi, Dirk A Loose, and Massimo Vaghi.Hemangiomas and vascular malformations:
an atlas of diagnosis and treatment. springer, 2015.
[3] Nikolaos Mouchtouris et al. “Management of cerebral cavernous malformations: from diagnosis to treatment”. In:The Scientific World Journal2015 (2015).
[4] Giannantonio Spena et al. “Preoperative and intraoperative brain mapping for the resection of eloquent-area tumors. A prospective analysis of methodology, correlation, and usefulness based on clinical outcomes”. In:Acta neurochirurgica152.11 (2010), pp. 1835–1846.
[5] Cristina Rosazza et al. “Preoperative mapping of the sensorimotor cortex: comparative assessment of task-based and resting-state FMRI”. In:PLoS One9.6 (2014).
[6] Han Lv et al. “Resting-state functional MRI: everything that nonexperts have always wanted to know”. In:American Journal of Neuroradiology39.8 (2018), pp. 1390–1399.
[7] Kathleen A Garrison et al. “The (in) stability of functional brain network measures across thresh-olds”. In:Neuroimage118 (2015), pp. 651–661.
[8] Richard S Snell.Clinical neuroanatomy. Lippincott Williams & Wilkins, 2010.
[9] Stephen G Waxman.Clinical neuroanatomy. McGraw Hill, 2010.
[10] John Harry Martin, Michael E Leonard, and Howard Radzyner. Neuroanatomy: text and atlas.
Elsevier New York, 1989.
[11] Virginia Stark-Vance and Mary Louise Dubay. 100 Questions & answers about brain tumors.
Jones & Bartlett Learning, 2010.
[12] Kevin Y Wang, Oluwatoyin R Idowu, and Doris DM Lin. “Radiology and imaging for cavernous malformations”. In:Handbook of clinical neurology. Vol. 143. Elsevier, 2017, pp. 249–266.
[13] Hesheng Liu et al. “Task-free presurgical mapping using functional magnetic resonance imaging intrinsic activity”. In:Journal of neurosurgery111.4 (2009), pp. 746–754.
[14] Stefan Kl¨oppel and Christian B¨uchel. “Alternatives to the Wada test: a critical view of functional magnetic resonance imaging in preoperative use”. In:Current Opinion in Neurology18.4 (2005), pp. 418–423.
[15] Michael G Hart et al. “Graph theory analysis of complex brain networks: new concepts in brain mapping applied to neurosurgery”. In:Journal of neurosurgery124.6 (2016), pp. 1665–1678.
[16] Elizabeth W Pang and OC Snead Iii. “From structure to circuits: the contribution of MEG connec-tivity studies to functional neurosurgery”. In:Frontiers in neuroanatomy10 (2016), p. 67.
[17] Sylvain Baillet. “Magnetoencephalography for brain electrophysiology and imaging”. In:Nature neuroscience20.3 (2017), p. 327.
[18] Andrei Irimia, John Darrell Van Horn, and Eric Halgren. “Source cancellation profiles of elec-troencephalography and magnetoencephalography”. In:Neuroimage59.3 (2012), pp. 2464–2474.
[19] A Hillebrand and GR Barnes. “A quantitative assessment of the sensitivity of whole-head MEG to activity in the adult human cortex”. In:Neuroimage16.3 (2002), pp. 638–650.
[20] Scott A Huettel, Allen W Song, Gregory McCarthy, et al.Functional magnetic resonance imaging.
Vol. 1. Sinauer Associates Sunderland, MA, 2004.
[21] Marie T Banich and Rebecca J Compton. Cognitive neuroscience. Cambridge University Press, 2018.
[22] Jerrold T Bushberg et al.The essential physics of medical imaging. Lippincott Williams & Wilkins, 2003.
[23] Christoph Stippich.Clinical functional MRI: presurgical functional neuroimaging. Springer, 2015.
[24] Paulo Branco et al. “Resting-state functional magnetic resonance imaging for language preopera-tive planning”. In:Frontiers in human neuroscience10 (2016), p. 11.
[25] C´esar Caballero-Gaudes and Richard C Reynolds. “Methods for cleaning the BOLD fMRI signal”.
In:Neuroimage154 (2017), pp. 128–149.
[26] John Ashburner et al. “SPM12 manual”. In: Wellcome Trust Centre for Neuroimaging, London, UK(2014), p. 2464.
[27] Theodore D Satterthwaite et al. “An improved framework for confound regression and filtering for control of motion artifact in the preprocessing of resting-state functional connectivity data”.
In:Neuroimage64 (2013), pp. 240–256.
[28] Yan Chao-Gan and Zang Yu-Feng. “DPARSF: a MATLAB toolbox for “pipeline” data analysis of resting-state fMRI”. In:Frontiers in systems neuroscience4 (2010).
[29] Michael G Hart, Stephen J Price, and John Suckling. “Connectome analysis for pre-operative brain mapping in neurosurgery”. In:British journal of neurosurgery30.5 (2016), pp. 506–517.
[30] Mikail Rubinov and Olaf Sporns. “Complex network measures of brain connectivity: uses and interpretations”. In:Neuroimage52.3 (2010), pp. 1059–1069.
[31] Mite Mijalkov et al. “BRAPH: a graph theory software for the analysis of brain connectivity”. In:
PloS one12.8 (2017).
[32] Yongqiang Yu et al. “Small-world brain network and dynamic functional distribution in patients with subcortical vascular cognitive impairment”. In:PloS one10.7 (2015).
[33] Yifei Zhu et al. “Disrupted brain connectivity networks in acute ischemic stroke patients”. In:
Brain imaging and behavior11.2 (2017), pp. 444–453.
[34] Wouter De Baene, Geert-Jan M Rutten, and Margriet M Sitskoorn. “Cognitive functioning in glioma patients is related to functional connectivity measures of the non-tumoural hemisphere”.
In:European Journal of Neuroscience50.12 (2019), pp. 3921–3933.
[35] Linqiong Sang et al. “Progressively disrupted brain functional connectivity network in subcortical ischemic vascular cognitive impairment patients”. In:Frontiers in neurology9 (2018), p. 94.
[36] Mite Mijalkov et al.BRAPH 1.0.0. Braph.org, 2016.
[37] Ben J Harrison et al. “Consistency and functional specialization in the default mode brain net-work”. In:Proceedings of the National Academy of Sciences105.28 (2008), pp. 9781–9786.
[38] Ji Eun Park et al. “Alteration of long-distance functional connectivity and network topology in patients with supratentorial gliomas”. In:Neuroradiology58.3 (2016), pp. 311–320.
[39] Qingling Huang et al. “Disturbed small-world networks and neurocognitive function in frontal lobe low-grade glioma patients”. In:PLoS One9.4 (2014).
[40] Jinfang Wang et al. “The role of disturbed small-world networks in patients with white matter lesions and cognitive impairment revealed by resting state function magnetic resonance images (rs-fMRI)”. In:Medical science monitor: international medical journal of experimental and clin-ical research25 (2019), p. 341.
[41] JD Kruschwitz et al. “GraphVar: a user-friendly toolbox for comprehensive graph analyses of functional brain connectivity”. In:Journal of neuroscience methods245 (2015), pp. 107–115.
[42] Jinhui Wang et al. “GRETNA: a graph theoretical network analysis toolbox for imaging connec-tomics”. In:Frontiers in human neuroscience9 (2015), p. 386.
[43] Misun Yoon and Hae-Jeong Park. “Multimodal brain NETwork analysis Toolbox (MNET) Soft-ware User’s Manual Version 1.1”. In: (2013).
[44] Alex Fornito, Andrew Zalesky, and Edward T Bullmore. “Network scaling effects in graph ana-lytic studies of human resting-state FMRI data”. In:Frontiers in systems neuroscience4 (2010), p. 22.
[45] JP Verma and Abdel-Salam G Abdel-Salam. Testing statistical assumptions in research. John Wiley & Sons, 2019, p. 67.
[46] Thomas W MacFarland and Jan M Yates. “Wilcoxon matched-pairs signed-ranks test”. In: Intro-duction to Nonparametric statistics for the biological sciences using R. Springer, 2016, pp. 133–
175.
[47] Henry Hsu and Peter A Lachenbruch. “Paired t test”. In:Encyclopedia of Biostatistics6 (2005).
Nodal results
This appendix intends to present the nodal graph theory results and represent them in the correspond-ing hemispheric-brain view for both lesioned and non lesioned networks, when thresholdcorrespond-ing them at a range of proportional thresholds from 20% to 40% with increments of 5%. It also aims to exhibit their corresponding nodal statistical results.
A.1 Graph Theory Analysis
In this section the nodal results from graph theory analysis will be exploit, including the mean and median values for both the global efficiency and clustering coefficient graph measures. Further, their mean values will be represented in a sagittal brain-view for both networks thresholded at each propor-tional threshold.
Proportionalthreshold20%25%30%35%40%RegionsLesionedNonlesionedLesionedNonlesionedLesionedNonlesionedLesionedNonlesionedLesionedNonlesionedPreCG0.539±0.1020.539±0.0960.598±0.0980.603±0.0840.658±0.0780.646±0.0900.704±0.0690.682±0.0820.740±0.0790.718±0.077SFGdor0.386±0.2270.574±0.0700.452±0.2180.621±0.0750.484±0.2310.660±0.0800.573±0.1520.719±0.0820.615±0.1520.752±0.089ORBsup0.387±0.2080.353±0.1930.436±0.2350.394±0.2130.471±0.2550.424±0.2270.557±0.1580.453±0.2360.579±0.1540.482±0.244MFG0.403±0.2600.539±0.0730.519±0.1600.586±0.0790.563±0.1540.631±0.0680.610±0.1330.668±0.0580.660±0.1150.701±0.056ORBmid0.335±0.2010.392±0.0980.383±0.2010.436±0.1130.431±0.2140.476±0.1150.465±0.2240.522±0.1020.505±0.2390.557±0.101IFGoperc0.434±0.1400.529±0.0720.509±0.1000.582±0.0660.569±0.0820.633±0.0530.620±0.0780.674±0.0480.652±0.0830.709±0.058IFGtriang0.469±0.1290.489±0.1030.554±0.1110.532±0.1040.607±0.1010.568±0.0990.650±0.0840.607±0.1030.702±0.0820.651±0.099ORBinf0.391±0.2180.459±0.1440.485±0.1600.507±0.1370.537±0.1740.555±0.1470.594±0.1610.617±0.0970.656±0.1130.666±0.087ROL0.487±0.1140.526±0.0830.550±0.0840.584±0.0790.602±0.0830.623±0.0750.635±0.0960.664±0.0680.695±0.0870.710±0.056SMA0.430±0.1520.480±0.0970.502±0.1240.536±0.0950.556±0.1120.584±0.0860.614±0.1290.618±0.0820.664±0.1180.664±0.084OLF0.310±0.2350.370±0.1900.353±0.2680.446±0.1120.444±0.2300.488±0.1240.489±0.2480.530±0.1250.608±0.1150.565±0.111SFGmed0.366±0.2250.502±0.0640.440±0.2110.558±0.0790.478±0.2300.602±0.0730.580±0.1100.630±0.0710.628±0.0930.660±0.078ORBsupmed0.357±0.1570.378±0.1940.454±0.0950.438±0.1420.518±0.0740.503±0.1410.557±0.0780.542±0.1320.607±0.0880.586±0.113REC0.397±0.1110.398±0.0980.466±0.1290.434±0.1080.502±0.1420.493±0.0930.541±0.1340.536±0.1040.577±0.1330.566±0.098INS0.433±0.2090.553±0.0800.538±0.1140.602±0.0760.590±0.1070.641±0.0670.653±0.0910.685±0.0620.700±0.0790.719±0.072ACG0.439±0.2150.483±0.0870.548±0.1340.513±0.0890.587±0.1220.556±0.0880.630±0.1050.603±0.0980.674±0.0850.636±0.107DCG0.457±0.2240.532±0.2420.593±0.0970.643±0.1010.652±0.0830.687±0.1070.710±0.0860.726±0.1020.769±0.0850.756±0.096PCG0.456±0.0780.465±0.2100.530±0.0830.517±0.2340.569±0.0850.612±0.0980.617±0.0780.664±0.0920.648±0.0750.696±0.074HIP0.340±0.1970.488±0.0420.438±0.1410.551±0.0460.497±0.1050.596±0.0380.557±0.1130.641±0.0350.595±0.1030.686±0.036PHG0.316±0.1830.452±0.0910.400±0.1090.501±0.1030.470±0.1100.545±0.1070.509±0.1180.586±0.1090.574±0.1170.633±0.103AMYG0.326±0.2160.357±0.1650.391±0.2140.405±0.1870.434±0.2230.437±0.2040.481±0.2350.507±0.1110.527±0.2490.554±0.109CAL0.610±0.0500.624±0.0430.685±0.0430.672±0.0440.722±0.0470.711±0.0370.761±0.0470.756±0.0300.786±0.0390.794±0.023CUN0.558±0.0700.563±0.0690.637±0.0210.616±0.0530.693±0.0300.655±0.0620.746±0.0340.701±0.0590.788±0.0300.740±0.055LING0.565±0.1280.641±0.0580.625±0.1060.694±0.0610.682±0.0810.739±0.0530.731±0.0630.770±0.0440.772±0.0690.796±0.039SOG0.483±0.2050.527±0.1290.582±0.1600.567±0.1190.628±0.1460.611±0.1360.681±0.1290.651±0.1320.709±0.1190.688±0.126MOG0.502±0.2130.548±0.0870.567±0.1800.606±0.0720.621±0.1830.651±0.0850.700±0.1180.687±0.0710.749±0.1080.719±0.073IOG0.533±0.0730.563±0.0540.610±0.0510.620±0.0520.639±0.0510.663±0.0550.681±0.0500.700±0.0500.720±0.0490.730±0.045FFG0.595±0.0510.586±0.0820.651±0.0400.639±0.0770.695±0.0340.686±0.0680.748±0.0430.728±0.0770.778±0.0370.769±0.092PoCG0.461±0.2080.534±0.0840.574±0.1040.587±0.0750.617±0.1050.649±0.0760.668±0.0840.688±0.0730.699±0.0790.719±0.072SPG0.540±0.1090.548±0.0730.599±0.1120.589±0.0680.639±0.1120.640±0.0640.686±0.0960.679±0.0720.722±0.1080.712±0.070IPL0.508±0.0980.510±0.0680.578±0.0750.584±0.0670.630±0.0590.635±0.0590.669±0.0770.667±0.0640.706±0.0800.703±0.069SMG0.489±0.0540.562±0.0450.566±0.0650.627±0.0450.624±0.0580.666±0.0590.669±0.0590.704±0.0660.720±0.0530.740±0.049ANG0.452±0.0640.541±0.0930.508±0.0710.592±0.0870.562±0.0690.631±0.1010.601±0.0840.676±0.0960.648±0.0800.712±0.092PCUN0.562±0.0940.588±0.0550.630±0.0860.650±0.0470.679±0.0720.696±0.0450.733±0.0620.734±0.0390.772±0.0480.769±0.019PCL0.378±0.1910.441±0.1390.440±0.2200.521±0.1170.532±0.1320.566±0.1260.582±0.1180.594±0.1330.629±0.1240.630±0.132CAU0.323±0.2400.385±0.2090.379±0.2650.498±0.1320.481±0.2440.537±0.1320.526±0.2650.575±0.1460.623±0.1470.620±0.114PUT0.385±0.1380.512±0.0820.493±0.1100.573±0.0550.556±0.0970.624±0.0530.616±0.0970.667±0.0580.661±0.0960.703±0.056PAL0.247±0.1270.375±0.0860.332±0.1720.433±0.0800.361±0.1790.476±0.0980.400±0.2010.512±0.1030.516±0.1010.556±0.094THA0.220±0.2150.468±0.0670.262±0.2500.521±0.0850.348±0.2560.583±0.0900.499±0.1080.620±0.0880.555±0.0870.660±0.097HES0.311±0.1600.498±0.1010.376±0.1800.540±0.1100.433±0.2020.592±0.1210.489±0.2350.630±0.1120.540±0.2540.671±0.096STG0.517±0.1050.609±0.0580.585±0.0940.667±0.0590.644±0.0740.711±0.0610.699±0.0720.744±0.0610.740±0.0620.783±0.059TPOsup0.441±0.1490.409±0.2380.532±0.1040.508±0.1650.594±0.1070.555±0.1670.642±0.1100.623±0.1070.683±0.1120.669±0.105MTG0.610±0.0610.642±0.0610.674±0.0410.690±0.0600.726±0.0390.737±0.0550.762±0.0410.779±0.0460.803±0.0410.814±0.042TPOmid0.383±0.1000.254±0.2410.442±0.1150.333±0.2320.491±0.1220.360±0.2500.527±0.1270.387±0.2700.578±0.1180.424±0.292ITG0.577±0.0690.579±0.1100.642±0.0650.635±0.0910.692±0.0570.683±0.0980.746±0.0450.724±0.0980.781±0.0450.767±0.098
TableA.1:Globalefficiencymeanvaluesofeachnodeinbothlesionedandnonlesionedhemisphereswhosenetworkswerethresholdedatarangeofproportionalthresholdfrom20%to40%withincrementsof5%.RegionnamesareomittedbutreferredinTable3.2inMethods.Dataarepresentedasmean±standarddeviation.
Proportionalthreshold 20%25%30%35%40% RegionsLesionedNonlesionedLesionedNonlesionedLesionedNonlesionedLesionedNonlesionedLesionedNonlesioned PreCG0.5410.5410.6340.6090.6830.6330.7080.7090.7480.732 SFGdor0.5010.5520.5500.6130.5710.6600.6260.7130.6790.732 ORBsup0.4750.3610.5460.3780.5860.4560.6020.4970.6230.536 MFG0.4060.5330.4970.6000.5530.6420.5850.6870.6260.719 ORBmid0.3360.3710.3890.4040.4300.4640.4910.5250.5530.581 IFGoperc0.4180.5180.5140.5780.5800.6030.6530.6480.6790.686 IFGtriang0.5020.4500.6040.4950.6480.5530.6820.6210.7300.689 ORBinf0.5100.4800.5350.5420.6160.5860.6450.6280.6790.675 ROL0.4820.5400.5270.6010.5690.6640.5850.6970.6760.732 SMA0.4840.4900.5380.5510.5530.6010.6230.6260.6760.667 OLF0.3580.3930.4010.4390.4260.4950.5600.5530.6510.585 SFGmed0.4930.5100.5270.5380.5520.5980.5680.6010.5910.628 ORBsupmed0.3950.4720.4730.4940.5500.5510.5860.5790.6420.607 REC0.3540.4030.4190.4430.4500.4840.5000.5510.5410.577 INS0.4920.5850.5710.6380.6210.6580.6670.7130.7300.735 ACG0.5060.5040.6050.5330.6340.5710.6460.6350.6730.670 DCG0.5280.6090.6190.6640.6680.6940.7300.7450.7920.779 PCG0.4630.5380.5020.6160.5490.6640.5970.6970.6540.713 HIP0.3920.4940.4520.5660.5050.6130.5520.6500.6150.691 PHG0.3540.4890.3910.5410.4830.5850.5240.6490.5970.689 AMYG0.3120.4090.3750.4540.4500.5010.5330.5340.6040.556 CAL0.6070.6340.6750.6770.7090.7100.7480.7380.7700.794 CUN0.5940.5460.6380.6220.6920.6580.7470.7190.7860.754 LING0.5740.6470.6190.7120.6860.7420.7340.7660.8050.800 SOG0.5850.5560.6270.5920.6710.6640.7390.6940.7550.724 MOG0.5550.5360.6050.6000.6760.6530.7360.6860.7800.705 IOG0.5490.5890.6190.6540.6380.6800.6980.6890.7260.708 FFG0.6160.5950.6570.6390.7030.6920.7530.7360.7990.806 PoCG0.5500.5590.6260.6080.6600.6610.6860.6800.7170.694 SPG0.5660.5590.5970.6060.6510.6380.7040.6840.7230.702 IPL0.5070.5300.6030.5830.6470.6310.6900.6690.7300.694 SMG0.4840.5700.5930.6380.6380.6830.6670.7130.7110.751 ANG0.4210.5900.4940.6190.5600.6710.5640.7140.6380.743 PCUN0.5490.6050.6380.6580.6920.7050.7440.7400.7780.770 PCL0.4530.4470.5000.4950.5240.5490.5750.6040.6450.642 CAU0.3770.4550.5110.5080.5940.5700.6090.6220.6230.656 PUT0.3540.5180.4180.5740.5240.6360.5720.6670.6100.726 PAL0.2610.3550.3360.4060.3750.4350.4210.4850.4950.548 THA0.3300.4850.4040.5370.4570.6010.5080.6490.5440.713 HES0.3420.4560.4310.5080.4560.5610.4950.5860.5580.648 STG0.5170.6300.5880.6930.6600.7230.7240.7510.7500.795 TPOsup0.4710.4930.5250.5550.6480.6330.6750.6570.7330.721 MTG0.6340.6500.6790.7140.7260.7530.7740.8010.8140.820 TPOmid0.3640.3630.4310.4110.4480.4640.4810.4950.5310.557 ITG0.5490.6340.6340.6650.6820.7050.7330.7400.7580.787 TableA.2:Globalefficiencymedianvaluesofeachnodeinbothlesionedandnonlesionedhemisphereswhosenetworkswerethresholdedatarangeofproportionalthresholdfrom20%to40% withincrementsof5%.RegionnamesareomittedbutreferredinTable3.2inMethods.
(a)Lesioned (b)Non lesioned
Figure A.1: Nodal global efficiency mean values obtained for lesioned (left) and non lesioned (right) networks when threshold-ing them at densities of 20% (top), 25% (middle) and 30% (bottom). The results are visualized in a sagittal hemispheric-brain view. Each one of the 45 hemispheric regions is displayed as a sphere along with the corresponding label of its name. High-er/lower values are represented by larger/smaller and darker/clearer spheres, respectively. Region names are omitted but referred in Table 3.2 in Methods.
(a)Lesioned (b)Non lesioned
Figure A.2: Nodal global efficiency mean values obtained for lesioned (left) and non lesioned (right) networks when thresh-olding them at densities of 35% (top) and 40% (bottom). The results are visualized in a sagittal hemispheric-brain view. Each one of the 45 hemispheric regions is displayed as a sphere along with the corresponding label of its name. Higher/lower values are represented by larger/smaller and darker/clearer spheres, respectively. Region names are omitted but referred in Table 3.2 in Methods.
Proportionalthreshold20%25%30%35%40%RegionsLesionedNonlesionedLesionedNonlesionedLesionedNonlesionedLesionedNonlesionedLesionedNonlesionedPreCG0.619±0.2160.522±0.1140.651±0.2040.577±0.1200.641±0.1650.644±0.1370.656±0.1110.686±0.1160.705±0.0910.697±0.117SFGdor0.530±0.2800.605±0.0940.554±0.2890.656±0.1110.560±0.2870.704±0.1190.717±0.1830.679±0.0920.749±0.1390.702±0.096ORBsup0.440±0.3420.594±0.3100.649±0.3130.561±0.2710.656±0.3140.593±0.2740.740±0.1610.656±0.2980.788±0.1310.671±0.311MFG0.499±0.2740.517±0.1640.489±0.2480.565±0.0670.517±0.2760.639±0.0690.694±0.2020.644±0.1060.600±0.2050.678±0.108ORBmid0.295±0.3750.633±0.3480.407±0.3500.753±0.1960.522±0.3320.712±0.1600.497±0.2710.712±0.1570.579±0.2830.782±0.101IFGoperc0.381±0.2890.641±0.1310.613±0.0910.683±0.1160.630±0.1620.704±0.0750.659±0.1520.713±0.0690.681±0.1360.738±0.076IFGtriang0.614±0.2240.744±0.2160.633±0.2080.776±0.1860.718±0.1480.789±0.1560.723±0.1060.813±0.1540.762±0.1020.790±0.110ORBinf0.434±0.3020.455±0.2200.442±0.3030.517±0.1050.468±0.3250.544±0.0930.728±0.1570.572±0.0760.652±0.1660.602±0.060ROL0.711±0.1820.609±0.0990.681±0.2140.655±0.0830.685±0.1610.709±0.0550.732±0.1440.718±0.0960.769±0.1220.727±0.072SMA0.692±0.3310.532±0.2650.803±0.2350.577±0.1400.746±0.1570.693±0.1080.759±0.1310.687±0.1200.781±0.0540.712±0.085OLF0.446±0.4220.615±0.3670.460±0.4350.663±0.3700.634±0.4390.715±0.3400.608±0.4190.829±0.1160.685±0.3380.845±0.089SFGmed0.493±0.3190.586±0.1810.558±0.3100.615±0.1620.575±0.2810.651±0.1330.617±0.1720.710±0.1170.647±0.1860.735±0.132ORBsupmed0.478±0.3930.655±0.3590.595±0.2810.677±0.3440.659±0.2230.661±0.3180.652±0.1870.739±0.2030.704±0.1570.793±0.111REC0.736±0.2890.813±0.2500.765±0.2050.728±0.2840.842±0.1670.759±0.2340.828±0.1260.793±0.1910.843±0.1340.818±0.151INS0.440±0.3230.538±0.0990.601±0.1370.623±0.1140.624±0.2220.659±0.0970.695±0.1270.664±0.0720.705±0.1340.693±0.057ACG0.587±0.3260.733±0.1790.609±0.3090.760±0.1980.664±0.2580.762±0.1710.757±0.1320.745±0.1570.765±0.1080.742±0.167DCG0.591±0.3230.516±0.2670.601±0.1850.612±0.1390.700±0.0990.624±0.1180.701±0.1030.668±0.1250.698±0.0910.716±0.139PCG0.686±0.3430.592±0.3120.611±0.3090.615±0.3090.801±0.1240.624±0.3040.766±0.1340.639±0.2990.792±0.1550.787±0.103HIP0.297±0.2240.606±0.1530.362±0.2570.632±0.1670.550±0.1500.667±0.1510.634±0.1180.698±0.1420.664±0.0780.705±0.129PHG0.573±0.4550.638±0.3140.583±0.4620.647±0.3190.565±0.3150.657±0.3240.662±0.3130.644±0.3150.668±0.1810.701±0.121AMYG0.485±0.4630.473±0.3740.579±0.4120.589±0.3360.727±0.3410.634±0.3530.708±0.3200.643±0.3570.678±0.3220.852±0.132CAL0.609±0.1360.606±0.0890.621±0.1580.649±0.0810.670±0.1650.675±0.0560.693±0.1520.684±0.0700.721±0.1400.684±0.047CUN0.694±0.1460.630±0.0900.695±0.1080.658±0.1020.714±0.1090.686±0.1160.712±0.0920.692±0.1180.720±0.0830.694±0.106LING0.670±0.2240.533±0.0850.716±0.2200.575±0.0960.721±0.1680.602±0.0890.723±0.1360.638±0.0690.722±0.1180.672±0.049SOG0.641±0.3170.527±0.2630.587±0.2210.584±0.2070.654±0.1860.598±0.1980.675±0.1250.656±0.1060.698±0.1410.661±0.125MOG0.525±0.2890.691±0.1380.561±0.2870.678±0.0970.550±0.2770.715±0.0930.614±0.1620.717±0.1270.630±0.1380.720±0.115IOG0.757±0.1370.677±0.0950.746±0.1280.701±0.0990.797±0.1060.723±0.0740.812±0.1020.751±0.0680.815±0.1080.765±0.054FFG0.621±0.1220.556±0.1230.641±0.1570.594±0.1160.664±0.1510.603±0.0980.692±0.1380.625±0.1010.708±0.1290.645±0.073PoCG0.628±0.3140.588±0.2420.653±0.3280.651±0.1180.726±0.1380.632±0.1050.712±0.1690.678±0.1160.769±0.1320.715±0.104SPG0.666±0.1180.637±0.0980.731±0.1240.665±0.1060.775±0.1210.699±0.0790.764±0.1140.721±0.0750.778±0.1310.729±0.099IPL0.632±0.1620.724±0.2260.647±0.1310.722±0.1650.659±0.1250.705±0.1240.705±0.1170.753±0.0880.728±0.1140.755±0.084SMG0.745±0.1270.674±0.0810.743±0.1270.683±0.0820.757±0.1340.718±0.1000.796±0.1110.737±0.1000.791±0.1020.768±0.096ANG0.877±0.1210.624±0.1650.778±0.1020.711±0.0850.786±0.1090.743±0.0770.801±0.0760.719±0.0920.784±0.0670.731±0.100PCUN0.634±0.2040.575±0.1010.621±0.1180.595±0.0850.615±0.1190.625±0.0760.651±0.0940.637±0.0780.671±0.0830.663±0.096PCL0.664±0.3630.403±0.3000.729±0.3290.613±0.3090.737±0.3310.780±0.1280.901±0.0650.811±0.1120.860±0.0750.839±0.130CAU0.461±0.4590.549±0.3950.557±0.3960.692±0.3210.531±0.3780.821±0.1030.673±0.3250.804±0.1440.675±0.3390.835±0.118PUT0.413±0.3970.544±0.1330.443±0.3460.636±0.0730.595±0.2070.684±0.0920.652±0.1530.669±0.1070.705±0.0900.705±0.095PAL0.414±0.5180.460±0.4630.356±0.4010.855±0.1840.447±0.4300.874±0.1660.492±0.4650.906±0.1310.869±0.1350.866±0.107THA0.210±0.3600.619±0.3080.277±0.3530.751±0.1760.402±0.3890.699±0.1350.485±0.3800.736±0.1180.688±0.2180.759±0.123HES0.775±0.3690.651±0.2700.736±0.3420.721±0.1500.757±0.3440.724±0.1350.700±0.3210.759±0.1330.703±0.3190.779±0.139STG0.582±0.1280.540±0.0630.603±0.1340.571±0.0600.632±0.1550.603±0.0720.655±0.1400.649±0.0840.698±0.1450.646±0.105TPOsup0.660±0.2280.361±0.2600.596±0.1030.382±0.2680.637±0.1090.639±0.1730.680±0.0760.574±0.1110.703±0.0880.641±0.091MTG0.540±0.1390.483±0.0360.589±0.1260.524±0.0520.624±0.1140.558±0.0730.660±0.1290.577±0.0790.674±0.1190.605±0.056TPOmid0.406±0.4650.236±0.3700.466±0.4370.326±0.3430.619±0.3960.472±0.4090.671±0.3570.506±0.4010.731±0.1580.491±0.353ITG0.589±0.1320.528±0.0670.627±0.1120.558±0.1180.665±0.1150.598±0.0870.673±0.1420.607±0.0810.696±0.1280.620±0.064
TableA.3:Clusteringcoefficientmeanvaluesofeachnodeinbothlesionedandnonlesionedhemisphereswhosenetworkswerethresholdedatarangeofproportionalthresholdfrom20%to40%withincrementsof5%.RegionnamesareomittedbutreferredinTable3.2inMethods.Dataarepresentedasmean±standarddeviation.
Proportionalthreshold 20%25%30%35%40% RegionsLesionedNonlesionedLesionedNonlesionedLesionedNonlesionedLesionedNonlesionedLesionedNonlesioned PreCG0.6410.5100.6150.5500.6320.6430.6320.7130.6770.754 SFGdor0.5280.6160.6670.6970.5450.7440.7520.7110.7100.744 ORBsup0.4750.6210.7580.6380.7540.6670.7810.7620.8080.736 MFG0.5500.5230.5710.5490.5870.6130.6280.6280.6360.664 ORBmid0.0000.6670.3330.7270.5270.6760.5000.6910.6570.736 IFGoperc0.4720.6430.6000.7220.6680.6860.6500.7200.6920.723 IFGtriang0.6360.6060.7000.7620.7020.7650.7330.7950.7870.775 ORBinf0.6000.5000.6080.5270.6580.5150.6980.5280.7170.589 ROL0.7780.6030.6940.6670.6820.6880.7580.7060.7450.722 SMA0.7620.5240.8610.6070.7610.7210.7430.7420.7780.750 OLF0.6670.6360.6890.7360.8670.8000.8100.8100.8240.833 SFGmed0.4290.6490.6000.6250.6440.6710.6280.7030.6830.744 ORBsupmed0.4670.7560.6000.8180.7580.7210.7250.7790.7120.800 REC0.8420.9340.7860.8600.8400.8570.8720.9050.8770.870 INS0.5270.5670.6220.6100.6490.6540.6950.6480.7080.706 ACG0.6670.8140.6230.8440.6490.8180.7130.7970.7780.768 DCG0.6220.5860.5690.6670.7380.6240.6650.6620.7040.706 PCG0.7880.6000.6670.6800.8100.7240.7360.7760.7940.818 HIP0.3810.6670.4640.7090.5560.6840.6030.7380.6580.788 PHG0.7330.6970.7730.7520.6000.7520.7910.7140.7500.741 AMYG0.7080.6000.7140.5780.8670.7500.8180.7600.7820.877 CAL0.6810.6450.6550.6530.6740.6940.6980.6920.7600.690 CUN0.7080.6410.6610.6280.6970.6400.6640.6830.7100.684 LING0.6630.5080.7750.5700.7540.6020.7680.6560.7640.658 SOG0.6430.6030.5530.6310.6550.6270.6670.6390.6690.665 MOG0.6210.6190.5930.6840.6120.7140.5890.6880.5890.695 IOG0.8030.6670.8010.6970.8010.7260.7730.7550.7550.778 FFG0.6670.5700.7210.6590.6820.5670.6610.5730.6780.642 PoCG0.6740.6240.7720.6780.7110.6260.6790.6880.7140.715 SPG0.6260.6490.7140.6360.7450.6920.7350.6970.7640.708 IPL0.6030.8180.6100.7750.6290.7120.6800.7580.7250.710 SMG0.7080.6910.7420.6840.7090.7220.8000.7510.8060.764 ANG0.8550.6190.7560.7390.7430.7860.8050.7100.7940.737 PCUN0.5890.5810.6270.6020.6340.6430.6530.6390.6400.691 PCL0.8000.5640.8100.6440.8500.7830.9090.7920.8430.800 CAU0.5110.6490.6830.7830.6670.8000.7070.8300.7520.834 PUT0.6000.5240.5640.6260.5760.7310.6540.6480.7190.699 PAL0.0000.5000.1671.0000.6001.0000.8001.0000.9090.857 THA0.0000.6940.0000.7310.5090.6670.6360.6740.6430.778 HES1.0000.6140.8060.6670.8330.7010.8000.7440.7790.727 STG0.6100.5710.6160.5710.6670.6120.6830.6210.7130.653 TPOsup0.6810.4580.6000.4680.6670.5710.6720.5770.6610.635 MTG0.5140.4710.6090.5290.5950.5610.6130.5920.6640.591 TPOmid0.1670.0000.3330.3210.6730.5000.7350.6110.7000.564 ITG0.6340.5200.5810.5690.6620.5710.6540.5780.7020.607 TableA.4:Clusteringcoefficientmedianvaluesofeachnodeinbothlesionedandnonlesionedhemisphereswhosenetworkswerethresholdedatarangeofproportionalthresholdfrom20%to 40%withincrementsof5%.RegionnamesareomittedbutreferredinTable3.2inMethods.
(a)Lesioned
(b)Non lesioned
Figure A.3: Nodal clustering coefficient mean values obtained for lesioned (left) and non lesioned (right) networks when thresh-olding them at densities of 20% (top), 25% (middle) and 30% (bottom). The results are visualized in a sagittal hemispheric-brain view. Each one of the 45 hemispheric regions is displayed as a sphere along with the corresponding label of its name.
Higher/lower values are represented by larger/smaller and darker/clearer spheres, respectively. Region names are omitted but referred in Table 3.2 in Methods.
(a)Lesioned (b)Non lesioned
Figure A.4: Nodal clustering coefficient mean values obtained for lesioned (left) and non lesioned (right) networks when thresholding them at densities of 35% (top) and 40% (bottom). The results are visualized in a sagittal hemispheric-brain view.
Each one of the 45 hemispheric regions is displayed as a sphere along with the corresponding label of its name. Higher/lower values are represented by larger/smaller and darker/clearer spheres, respectively. Region names are omitted but referred in Table 3.2 in Methods.