Alahmadi Insights Imaging (2021) 12:47 https://doi.org/10.1186/s13244-021-00993-9 Insights into Imaging

ORIGINAL ARTICLE Open Access Investigating the sub‑regions of the superior parietal cortex using functional magnetic resonance imaging connectivity Adnan A. S. Alahmadi*

Abstract Objectives: Traditionally, the superior parietal lobule (SPL) is usually investigated as one region of interest, particu- larly in functional magnetic resonance imaging (fMRI) studies. However, cytoarchitectonic analysis has shown that the SPL has a complex, heterogeneous topology that comprises more than seven sub-regions. Since previous studies have shown how the SPL is signifcantly involved in diferent neurological functions—such as visuomotor, cognitive, sensory, higher order, and attention—this study aims to investigate whether these cytoarchitectur- ally diferent sub-regions have diferent functional connectivity to diferent functional brain networks. Methods: This study examined 198 healthy subjects using resting-state fMRI and investigated the functional con- nectivity of seven sub-regions of the SPL to eight regional functional networks. Results: The fndings showed that most of the seven sub-regions were functionally connected to these targeted networks and that there are diferences between these sub-regions and their functional connectivity patterns. The most consistent functional connectivity was observed with the visual and attention networks. There were also clear functional diferences between Brodmann area (BA) 5 and BA7. BA5, with its three sub-regions, had strong functional connectivity to both the sensorimotor and salience networks. Conclusion: These fndings have enhanced our understanding of the functional organisations of the complexity of the SPL and its varied topology and also provide clear evidence of the functional patterns and involvements of the SPL in major brain functions. Keywords: FMRI, Functional connectivity, Superior parietal cortex

Key points • Te most consistent functional connectivity was observed with the visual and attention networks. • Te SPL has a complex, heterogeneous topology that • Tere were also clear functional diferences between comprises more than seven sub-regions Brodmann area (BA) 5 and BA7. • Te functional connectivity of these sub-regions to diferent functional brain networks was investigated. • Tere are similarities and diferences between these Introduction sub-regions and their functional connectivity pat- Te superior parietal lobule (SPL) plays an important role terns. in diferent brain functions including visuomotor, cogni- tive, sensory, higher order, working memory and atten- tional [1–10]. Most of these fndings were investigated *Correspondence: [email protected]; [email protected] using functional magnetic resonance imaging (fMRI) Department of Diagnostic Radiology, College of Applied Medical Science, King Abdulaziz University , Jeddah, Saudi Arabia

© The Author(s) 2021. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creat​ iveco​ mmons.​ org/​ licen​ ses/​ by/4.​ 0/​ . Alahmadi Insights Imaging (2021) 12:47 Page 2 of 12

based on specifc experimental tasks, and the fndings “Te 1,000 Project data-sharing efort reported relate to the SPL as a single region. will provide the research community with open access Another powerful and useful technique that aids the to datasets contributed by labs around the world. Data- understanding of the brain’s functional networking sets provided to the 1,000 Connectomes Project are to mechanism is resting-state fMRI (rsfMRI), which helps be de-identifed prior to deposition of the data with the to determine the functional integrations of regions of project (i.e. removal of any personal identifying informa- the brain during rest [11]. rsfMRI is also a powerful tool tion from header/support fles). Upon arrival, datasets regarding the structural understanding of cortical organi- are automatically organised, and header fles are replaced sations, as emerging evidence from studies have shown with novel header fles to guarantee that any identifying that these functional, spontaneous low-frequency signals personal information within the header or supporting were signifcantly structurally correlated [12–14]. Evi- fles is removed. Prior to open-access sharing via web- dence also suggests that these spontaneous fuctuations based repository, all datasets will be further de-identifed observed during rsfMRI produced functional networks and anonymised by the removal of face information from that were continuously correlated with experimental the image to prevent any inappropriate reconstruction of tasks [13]. the image that could lead to the identifcation of a par- In addition, and in an efort to further understand ticipant. Furthermore, each participant’s dataset will the cortical organisation of the SPL, recent studies used be assigned a randomised fve-digit participant identi- cytoarchitecture analysis and probabilistic maps in ten fer, along with a site identifer (two letters which map post-mortem brains to analyse the SPL [15–17]. Te to the site providing the data). Te relationship between studies showed that the SPL was sub-divided into eight the anonymised code and the original subject ID will sub-regions: 5 Ci, 5 M, 5L, 7PC, 7A, 7P, 7 M and hIP3. be destroyed to ensure that the dataset will be truly Tree of these sub-regions are located in Brodmann (BA) anonymised. For each dataset, demographic information 5, and four of them are located in BA7. Each of these sub- provided via web-archive will be limited to (when avail- regions is diferent based on the receptor distribution able): age (18 and up), gender (male, female) and handed- patterns and the regional cytoarchitectonic properties. ness. Tis information will serve to facilitate more careful Te functional roles of each of these sub-regions have characterisation of the data, without entailing a risk of still not been fully investigated, and most of the previous violation of confdentiality. Datasets will only be included studies mentioned above claimed the involvement of SPL in the repository upon receipt of written expressed per- as a single cortical region. Te assumption is that since mission for usage of the dataset freely by the general pub- these sub-regions are diferent from an anatomical point lic, without limitation.” of view, and each has diferent cytoarchitectonic prop- Tis dataset was acquired using a typical rsfMRI pro- erties, their diferences can expect to be observed dur- tocol that is used by previous studies, see for review and ing functional connectivity to major network functions. references [18–20]. Te registered clinical name for this Terefore, this study aims to use a large cohort of sub- project is the Cambridge-Buckner dataset. All of the 198 jects to investigate the functional connectivity of each of subjects were scanned using a Siemens 3-T Trim Trio these seven sub-regions of the SPL with major functional scanner with the following parameters: T2* weighted identifed networks and to use rsfMRI to determine how Echo Planner Imaging (EPI) sequence with repetition these sub-regions difer based on functional connectivity. time (TR) 3 s; TE 30 ms; number of slices 47 inter- = = = 3 leaved axial slices; voxel size: 3.0 × 3.0 × 3.0 ­mm ; num- ber of time points (i.e. volumes) = 124 volumes (the frst 5 Methods volumes were discarded). T1-weighted MPRAGE images Subject recruitments and scanning with the following parameters: number of slices: 192; 3 A total of 198 healthy volunteer subjects were recruited matrix size 144 × 192; voxel size: 1.20 × 1.00 × 1.33 ­mm . for this study. Teir ages ranged from 18 to 30, and 123 of the subjects were female. 171 of the subjects were right Pre‑processing handed. Te data were collected and downloaded from Te pre-processing and statistical analyses were per- the Cambridge-Buckner dataset, which is part of the formed using CONN and Statistical Parametric Map- 1,000 Functional Connectomes Project (an open-access ping software (SPM12). Te pre-processing of the rsfMRI platform without any restrictions, https://​www.​nitrc.​org/​ imaging volumes included such typical pre-processing proje​cts/​fcon_​1000/) [12]. Tis is a non-restricted public steps as slice timing corrections, realignment of the func- dataset and is available to all. Te IRP statement provided tional volumes, normalisation of the functional volumes on the 1,000 Functional Connectomes Project website is to an MNI template using structural data, detection data shown on the following paragraph. outliers using the implanted artefact detection tools Alahmadi Insights Imaging (2021) 12:47 Page 3 of 12

(ART) in CONN [21], and smoothing the functional vol- Table 1 Targeted networks with the regions that comprise umes using an 8-mm kernel. In addition, in order to neu- them. Also shown the coordinates of these regions tralise the efects of artefacts and confounds (e.g. white Network Regions matter, CSF signals, and motion and scrubbing param- Default mode MPFC (1,55, 3) eters) from the BOLD signal, temporal processing with − LP (L) ( 39, 77,33) data denoising was also applied. − − LP (R) (47, 67,29) − PCC (1, 61,38) Selection of the regions of interest − Sensorimotor Lateral (L) ( 55, 12,29) Te seeds identifed in this study were the seven sub- − − Lateral (R) (56, 10,29) regions of the SPL, namely the BA5 sub-regions (5 Ci, − Superior (0, 31,67) 5 M, and 5L) and the BA7 sub-regions (7PC, 7A, 7P, − Visual Medial (2, 79,12) and 7 M) (Fig. 1). Tese regions were identifed using − Occipital (0, 93, 4) the cytoarchitectonic probability anatomy map, which − − Lateral (L) ( 37, 79,10) is guided by the cytoarchitectonic properties of these − − Lateral (R) (38, 72,13) regions based on the data from ten post-mortem brains − [15, 17, 22, 23]. Te target brain regions were major col- Salience ACC (0,22,35) Insula (L) ( 44,13,1) lections of functional networks, as defned in CONN − and shown in Table 1. In addition to the default mode, Insula (R) (47,14,0) RPFC (L) ( 32,45,27) the functional networks studied were attention, sensori- − motor, visual, salience, dorsal attention, frontoparietal, RPFC (R) (32,46,27) SMG (L) ( 60, 39,31) cerebellar, and language. Tis atlas of commonly used − − SMG (R) (62, 35,32) networks was defned based on an independent compo- − Dorsal Attention FEF (L) ( 27, 9,64) nent analysis (ICA) of the human connectivity project − − FEF (R) (30, 6,64) using 497 subjects. − IPS (L) ( 39, 43,52) − − Statistical analysis IPS (R) (39, 42,54) − Fronto-Parietal LPFC (L) ( 43,33,28) Statistical analysis was calculated at two levels. At − PPC (L) ( 46, 58,49) the frst level (subject-level), weighted general linear − − bivariate correlation models, including regions of inter- LPFC (R) (41,38,30) PPC (R) (52, 52,45) ests (ROI)-to-ROI connectivity matrices between the − Language IFG (L) ( 51,26,2) − IFG (R) (54,28,1) pSTG (L) ( 57, 47,15) − − pSTG (R) (59, 42,13) − Cerebellar Anterior (0, 63, 30) − − Posterior (0, 79, 32) − − Abbreviation: MPFC: medial prefrontal cortex; LP: lateral parietal; PCC: posterior ; ACC: anterior cingulate cortex; RPFC: rostrolateral prefrontal cortex; SMG: supramarginal gyrus; FEF: frontal eye felds; IPS: ; LPFC: lateral prefrontal cortex; PPC: posterior parietal cortex; LPFC: lateral prefrontal cortex; IFG: inferior frontal gyrus; pSTG: posterior superior temporal gyrus

pre-defned ROIs, were calculated for each subject. Tese connectivity matrices were defned as Fisher- transformed bivariate correlation coefcient between the pair of ROI timeseries. At the second level, func- tional connectivity measures were calculated and com- pared using group-level statistical analysis, such as T-tests and/or F-tests where appropriate, and identify- ing and comparing the rsfMRI networks connected to Fig. 1 Illustrations of the seven SPL source regions used in this study each of the sub-regions of the SPC at the group level. in the right hemisphere Te standard setting for the results is displayed using Alahmadi Insights Imaging (2021) 12:47 Page 4 of 12

a corrected false discovery rate (FDR) (p < 0.05) (mul- Results tivariate statistics parametric (MVPA) omnibus test) All of the sub-regions are connected to each other, and [21]. Here, cluster-level inferences are determined the functional connectivity between the seven sub- using functional network connectivity multivariate regions of the SPL is shown in Fig. 2. Figure 2 also shows parametric statistic inferences by considering groups that the strength of each of the connections is statistically or networks of related ROIs. Ten, the analysis is done diferent from each other. by analysing the entire set of connections between the Te study investigated the functional connectivity ROIs in terms of within- and between-network con- between the seven sub-regions of the SPC in the two nectivity [24]. Tis method efectively performs a mul- hemispheres of the brain to the eight targeted networks. tivariate parametric general linear model analysis for In general terms, the results showed that there are strong all connections. Te resultant map is of F-statistical connections (either positive or negative) between the tests for each pair of networks. Te FDR cluster level sub-regions and the targeted networks. Tere were simi- is defned as the expected proportion of false discover- larities and diferences among the connections, which ies among all pairs of networks with similar or larger are summarised in Figs. 3 and 4. Additionally, Fig. 5 sum- efects across the entire set of functional connectivity marises the connections by indicating whether there is network pairs [25]. One of the reasons for using the a positive connection, a negative connection, or no con- FDR over family wise error (FWE) is that the FDR is nection. Te key fndings are shown below. more sensitive in controlling peaks, with minimal cost of false positives [26]. For additional details, please Functional connectivity with the refer to the above papers or the CONN website (https://​ In general, the seven sub-regions of the SPL had posi- web.​conn-​toolb​ox.​org). tive connections with the regions defning the default mode network, particularly the lateral parietal (LP) and

Fig. 2 The functional connectivity among the sub-regions of the SPL is shown. These are the seven investigated sub-regions of the SPL in right and left hemispheres (14 sub-regions). The lines of the connections in red indicate positive connectivity, and these colours are proportional to statistical strength. Additionally, the T-bar is shown in the top right-hand corner Alahmadi Insights Imaging (2021) 12:47 Page 5 of 12

Fig. 3 The functional connectivity of the three SPL sub-regions of BA5 in both hemispheres is shown here. The lines of the connections in red indicate positive connectivity while blue indicates negative connectivity. The colours of the lines are proportional to statistical strength, and the T-bar is shown in the top right-hand corner Alahmadi Insights Imaging (2021) 12:47 Page 6 of 12

Fig. 4 The functional connectivity of the four SPL sub-regions of BA 7 in both hemispheres is shown here. The lines of the connections in red indicate positive connectivity while blue indicates negative connectivity. The colours of the lines are proportional to statistical strength, and the T-bar is shown in the top right-hand corner Alahmadi Insights Imaging (2021) 12:47 Page 7 of 12

Fig. 5 The functional connectivity of all the SPL sub-regions in both hemispheres is shown here, along with showing whether each sub-region is connected positively ( 1), is not connected (0) or is connected negatively (-1) with the eight targeted networks + posterior cingulate cortex (PCC). Negative connections sub-regions of the SPL located within BA5 had positive were also observed, particularly with the medial prefron- connections. In contrast, the right SPL 7 M and the left tal cortex (MPFC) and within BA5. Te results also indi- SPL 7PC had no connections at all. cated that there were no signifcant diferences between the two hemispheres as most of the connections to these Functional connectivity with the visual network sub-regions of the SPL were identical. Te sub-regions In general, most of the sub-regions of the SPL had posi- that had clear and distinct diferences from the other tive connections with the regions defning the visual net- sub-regions were the SPL 7 M, the SPL 7PC and the left work. Tis was highly constant within the sub-regions of SPL 5L. BA7. Within the sub-regions of BA5, there were negative connections with the SPL 5 M and SPL 5 Ci. Functional connectivity with the In general, most of the sub-regions of the SPL located Functional connectivity with the within BA7 had negative connections with the sub- Tere were clear diferences between the sub-regions of regions defning the sensorimotor network, while all the the SPL within BA7 and the regions defning the salience Alahmadi Insights Imaging (2021) 12:47 Page 8 of 12

network. Tese were either defned by positive, nega- posterior superior temporal gyrus (pSTG) of the lan- tive or no connections at all. While the sub-regions of guage network. In contrast, the only negative connec- SPL within BA5 had mostly positive connections, the tions in these two sub-regions were with the left IFG. sub-regions of the SPL within BA7 contained diferences Hemispheric diferences were not signifcantly observed between the two hemispheres when functional connec- among all of the sub-regions and it was mainly related to tivity with the salience network was investigated. Te the SPL 5 M and SPL 5 Ci. largest number of positive connections of the SPL BA7 sub-regions were seen with the rostrolateral prefrontal Functional connectivity with the cerebellar network cortex (RPFC) of the salience network, while the SPL 7A had negative connections with most of the regions mak- Tere is a clear diference between the sub-regions of ing up the salience network. SPL within BA7 and the regions defning the cerebel- lar network, with heterogeneous functional connectiv- Functional connectivity with the ity within the BA7 sub-regions being observed. Negative connections with the posterior cerebellum were constant In general, most of the sub-regions of the SPL had posi- with all the sub-regions of BA5, while negative connec- tive connections with the regions defning the dorsal tions were also observed with the SPL 7 M and the left attention network. Tis was highly constant within the SPL 7P. Te anterior cerebellum also showed negative sub-regions of BA5. Within the sub-regions of BA7, there connections with most of the sub-regions of BA5. Posi- were negative connections with the SPL 7 M and left SPL tive connections were mainly observed between the ante- 5P. SPL 7P only had positive connections with the intra- rior cerebellum and the SPL 7PC and SPL 7A. parietal sulcus (IPS) of the dorsal attention network, while the left SPL 7P had negative connections with the frontal eye felds (FEF) of the dorsal attention network. Discussion Te SPL 7 M had negative connections in both hemi- Tis study used a large dataset to investigate the func- spheres with the FEF and IPS of the dorsal attention net- tional connectivity of seven sub-regions of the SPL with work, while it had no connections with the IPS. eight regional network functions using rsfMRI. Tese seven sub-regions were shown to be diferent from each Functional connectivity with the other both structurally and cytoarchitecturally [15–17, In general, most of the sub-regions of the SPL had nega- 23, 27]. Te assumption was that since these sub-regions tive connections with the regions making up the fron- were cytoarchitecturally diferent, their functional con- toparietal network. Tese negative connections were nectivity could also be diferent. In general terms, the constant within the sub-regions of BA5. Te sub-regions fndings of this study showed that these sub-regions were of BA7 were heterogeneous in terms of their functional connected to the eight functional networks. Te func- connections with the regions defning the frontopari- tional connectivity was similar between some networks etal network, while SPL 7 M had negative connections. and diferent between other networks. Positive connections were only observed in SPL 7P with Te fndings showed the important functional role that the right lateral prefrontal cortex (LPFC) and the poste- the SPL plays in terms of its involvements in diferent rior parietal cortex (PPC) of the frontoparietal network. functional domains, particularly regarding the attentional Hemispheric diferences were observed in SPL 7PC, with and visual pathways. Here, the functional connectivity the right SPL 7PC having a negative connection with the was most strongly connected with the networks in most left PPC of the frontoparietal network. of the sub-regions of the SPL. Tis is in line with previ- ous fndings that showed a strong involvement of the Functional connectivity with the language network SPL with visual and attentional functions [2, 8, 28–30]. In general, most of the sub-regions of the SPL had nega- However, the connections to the visual and attention net- tive connections with the regions defning the language works were neither homogeneous nor consistent among network. Tese negative connections were mostly within the sub-regions of the SPL. the sub-regions of BA7. Te only region that had a posi- Te results of this study indicated that all of the sub- tive connection with the language network within BA7 regions of BA7 were signifcantly involved with the visual was the left SPL 7PC, which had a positive connection networks, whereas the sub-regions of BA5 were showing with the left inferior frontal gyrus (IFG). Te sub-regions some heterogeneous functional connections. For exam- of BA5 were partially heterogeneous in terms of their ple, there were no relationships between the sub-regions functional connections with the regions defning the of BA5 with the visual occipital part of the visual net- language network. Two sub-regions of BA5—SPL 5 M work. Tis visual occipital part was shown to be mainly and SPL 5 Ci—had positive connections with the right involved in the perception of visual shapes [13]. Alahmadi Insights Imaging (2021) 12:47 Page 9 of 12

Tis could indicate that not all of the SPL is involved brain functions controlled by the SPL. Te fact that in visual functions. SPL BA7 plays a greater role in vis- these sub-regions had more positive connections with ual processing, while SPL BA5 has limited involvements the LP and PCC of the default mode network could be in visual functions but no direct involvement in the per- because of their involvement in higher-order cognitive ception of visual shapes. When looking at the attentional and attentional tasks. network, which consisted of FEF and IPS, all of the sub- However, the results also showed that the default mode regions of SPL BA5 were directly involved in attentional network had negative connections with the MPFC. Tis functions to a greater extent than the sub-regions of SPL suggests that the default mode and the seven sub-regions BA7. Tere was heterogeneous connectivity within the of the SPL have some heterogeneous relationships and sub-regions of SPL BA7 where only the right SPL 7PC that the default mode network sub-regions should be had full connectivity with all of the sub-regions of the investigated as a heterogeneous network. Tis is in line attention network. Some studies have shown that the with studies that showed that the prefrontal cortical part, involvement of SPL in attentions was predominant in which had negative connections with most of the SPL the right hemisphere [9], which could indicate that this is sub-regions, and the PCC, which had positive connec- only true within SPL BA7 but not within SPL BA5. tions with most of the SPL sub-regions, are distinctly dif- In addition, the sensorimotor network involves areas ferent from each other [37]. that are related to motor and sensory functions, includ- Our study also showed that there were sub-regions of ing the pre- and post-central gyri as well as the supple- the SPL that were distinctly diferent from the other sub- mentary motor areas. Indeed, the SPL plays an important regions, namely the SPL 7 M, SPL 7PC and the left SPL role in sensorimotor tasks, particularly visuomotor tasks 5L. Recent studies showed that the SPL 7 M, SPL 7PC or tasks that require certain visuomotor attention [8]. and SPL 5L showed distinct functional activations or Te results of this study showed that there were dis- connectivity during diferent operational tasks [1, 38], tinct diferences between BA5 and BA7. Most of the which could explain their distinct patterns of connections sub-regions of BA5 had positive connectivity with the with the prefrontal part of the default mode network. sensorimotor networks, particularly the superior region, Moreover, the involvement of the SPL’s sub-regions whereas the BA7 sub-regions, except for SPL 7PC, had with the other remaining networks—namely the salience, either negative or no connections with the sensorimo- frontoparietal, language and cerebellar networks—were tor network. Previous studies have shown that BA5 has heterogeneous and were not strongly connected with all extensive involvement in sensorimotor functions, while of the sub-regions. In addition, anticorrelation or nega- BA7 has less involvement [31, 32]. Te results indicated tive connectivity was often observed between most of the that most of the sub-regions of BA5, apart from the right sub-regions of the SPL with these networks. However, the SPL 5 M, were indeed involved with the sensorimotor interpretation of these negative functional connections is network. beyond the scope of this study, particularly as there is still Te results also showed that the right SPL 7PC is the debate regarding the meaning of these anticorrelation only sub-region in BA7 that was involved with the senso- signals in functional connectivity [39]. Te results of this rimotor network, indicating that there are heterogeneous study could indicate that heterogeneous functional con- connections in the sub-regions of BA5 and BA7. How- nectivity is present and that the SPL should not be seen ever, these heterogeneous connections were only found or investigated as a whole region of interest. within one of the sub-regions. Two further observations are worth mentioning and Te default mode network is a network of interact- highlighting. SPL 7PC and SPL 7A were the two sub- ing brain regions that include the MPFC, the LP and regions found to be positively connected to the anterior the PCC. Tese interacting brain regions are related cerebellum, which is known to be involved in motor- to each other and usually have distinct patterns com- related functions [40–46]. Tis suggests that these two pared to other networks [33]. Typically, the default sub-regions could have a direct role in organising the mode network is related to intrinsic changes and it has pathways and functional involvements of the anterior greater remarkable diferences in neurological diseases cerebellum. Also, recent studies have found that SPL [34–36]. In our study, the seven sub-regions of the SPL 7PC was mainly involved in execution and motor func- had mostly positive connections with the regions defn- tions [1], which could explain the direct involvement of ing the default mode network, particularly with the LP this sub-region as a motor pathway with the anterior cer- and PCC of the default mode network. Tese fndings ebellum. Finally, this study also showed that the SPL BA5 indicate the importance of these sub-regions in their sub-regions were all involved and correlated positivity connections to the default mode network and indicate with the salience network. Te salience network is known how the default mode network may play a role in the to be involved in salience functions and attentions [47], Alahmadi Insights Imaging (2021) 12:47 Page 10 of 12

which could also suggest that BA5 has a role in driving Finally, future studies could investigate structural and the pathways of focused and salience functions. functional connectivity among those sub-regions to clar- ify the unexplained connectivity and better understand Methodological and future considerations the physiological organisations underlying the regions Despite the coherent fndings in this study and the clear with diferent networks. For example, it may be that diferences among the SPL sub-regions, there are some the internal interactions spontaneous patterns could be potential limitations that should be considered in future highly structured [1, 27, 57, 58]. Or these functional con- studies. Future studies could investigate the question of nections could not be correlated structurally, which may the diferences of the functional connectivity of these suggest mediations by indirect structural connections sub-regions using high-resolution data (i.e. 1 mm3). [58]. Also, future studies could investigate efective con- Tis may help to enhance the subtle diferences among nectivity and task-related activations of these sub-regions these sub-regions. Usually, such high-resolution data are [59]. Terefore, a multimodal approach to analyse these acquired using high feld scanner strength (i.e. 7 T). regions is needed. For example, a recent study aimed to In addition, a potential limitation of this study was the identify a convergent organisation of the SPL using struc- applied smoothing kernel. Although 8 mm3 is a default tural and functional connectivity [1]. Te researchers (and considered typical [48–51]) kernel for such voxel identifed fve sub-regions (two anterior and three pos- resolution, future studies may consider the efects of terior) in the SPL based on multimodal smoothing on the connectivity of these sub-regions. analyses. Tey were able to show functional-related dif- However, some recent studies have shown that smooth- ferences of the identifed regions of the brain that were ing kernels could have little efects, especially on rs-fMRI linked to vision, motor control, perception, memory and ROI to ROI analysis [52, 53]. Regardless of these stud- attention. Te fndings of their results are in line with the ies and since this study’s source regions of interest were fndings of our study in terms of the diferent functional adjacent, one could argue that the 8 mm3 smoothing ker- involvements of the SPL regions. One could therefore nel would blur those regions and afect the results. Tere- combine the usage of these cytoarchitecture identifed fore, a minimal smoothing kernel (or no smoothing) sub-regions and with structurally identifed approaches could be tested in future studies to compare and validate to further enhance our understanding of the physiologi- the results. cal organisations of these sub-regions. Another limitation of this study is the usage of an Terefore, it must also be acknowledged that interpret- open-access dataset. Te limitation is related to the infor- ing the specifc connectivity patterns of these sub-regions mation that could be used in the statistical method to is difcult to do. Tis is because, on the one hand, difer- investigate, for example, if gender and education level ent task experimental fMRI needs to be implemented to could be confounding factors that may have afected the test for the involvement of these sub-regions in diferent results. Previous studies have shown functional connec- brain functions. On the other hand, this study is one of tivity based on gender diferences [54]; however, such the frst that attempts to understand the cortical hetero- factors need careful data acquisition and planning before geneous functional connectivity of these cytoarchitectur- starting studies, and performing such was beyond the ally diferent sub-regions. aim of this study. In addition, handedness in this study However, the fndings of this study are important was shown to have no efect on the investigated results, because they provide clear evidence that in addition when used as a covariate in the GLM. Tis fnding is to the cytoarchitectural diferences among these seven likely because the number of right-handed subjects was SPL sub-regions, there are also functional connectivity large compared to left-handed subjects, and this study diferences that should also be considered. Investigat- did not aim to investigate the efect of handedness on the ing the similarities and diferences in functional con- results. Tus, future studies could design a study to inves- nectivity among these seven sub-regions of the SPL tigate such efects, which probably could afect the func- with important targeted functional brain networks tional connectivity of SPL 7A and 5 M with other brain should highlight the importance of using more speci- regions. fed detailed anatomical atlases to investigate these In addition, one of this study’s limitations is the ana- cortical regions in more specifc ways. It could also lysed dataset’s parameters. For example, this study’s help future studies to focus on the impacts of diferent number of time points is not large. Longer scans (and neurodegenerative diseases, and how the pathologies more time points) can increase resting state reliability of such diseases afect the functional connectivity of and improve sensitivity [55, 56]. Terefore, future studies these sub-regions. Te fndings of this study could also could consider better quality datasets to check the reli- enhance our understanding of the functional organi- ability of this study’s fndings. sations of the complexity of the SPL and its varied Alahmadi Insights Imaging (2021) 12:47 Page 11 of 12

topology and could also assist in reviewing the applied Received: 13 January 2021 Accepted: 23 March 2021 specifc task functional activation involvements of these sub-regions.

References 1. Wang J et al (2015) Convergent functional architecture of the superior Conclusion parietal lobule unraveled with multimodal neuroimaging approaches. Tis study has shown the similarities and diferences in Hum Brain Mapp 36(1):238–257 2. Culham JC, Valyear KF (2006) Human parietal cortex in action. Curr Opin functional connectivity between the seven cytoarchi- Neurobiol 16(2):205–212 tecturally diferent sub-regions of the SPL with difer- 3. Weiss PH et al (2003) Are action and perception in near and far space ent functional brain networks. Te study has shown additive or interactive factors? Neuroimage 18(4):837–846 4. Vingerhoets G et al (2002) Motor imagery in mental rotation: an fMRI how each of the SPL sub-regions plays an important study. Neuroimage 17(3):1623–1633 role in visual and attentional functions. Te study 5. Zago L, Tzourio-Mazoyer N (2002) Distinguishing visuospatial working has also shown the various involvements of the sub- memory and complex mental calculation areas within the parietal lobes. Neurosci Lett 331(1):45–49 regions of BA7 and BA5 in other brain functions. Te 6. Wenderoth N et al (2004) Parieto-premotor areas mediate direc- heterogeneity of the functional connectivity of these tional interference during bimanual movements. Cereb Cortex sub-regions has also been proven by this study, which 14(10):1153–1163 7. Bray S et al (2013) Structural connectivity of visuotopic intraparietal suggests that the SPL has diferent complex structural sulcus. Neuroimage 82:137–145 and functional topological organisations that should 8. Corbetta M et al (1995) Superior parietal cortex activation dur- always be considered when investigating the physiology ing spatial attention shifts and visual feature conjunction. Science 270(5237):802–805 of this important cortical region, or in neurological dis- 9. Coull J, Frith C (1998) Diferential activation of right superior parietal ease applications. cortex and intraparietal sulcus by spatial and nonspatial attention. Neuro- image 8(2):176–187 10. Lloyd D, Morrison I, Roberts N (2006) Role for human posterior parietal Abbreviations cortex in visual processing of aversive objects in peripersonal space. J ART​: Artefact detection tools; BA: Brodmann area; FDR: False discovery rate; Neurophysiol 95(1):205–214 fMRI: Functional magnetic resonance imaging; IPS: Intraparietal sulcus; LP: 11. Biswal B et al (1995) Functional connectivity in the motor cortex of rest- Lateral parietal; LPFC: Lateral prefrontal cortex; MPFC: Medial prefrontal cortex; ing human brain using echo-planar MRI. Magn Reson Med 34(4):537–541 PCC: Posterior cingulate cortex; ROI: Regions of interests; RPFC: Rostrolateral 12. Biswal BB et al (2010) Toward discovery science of human brain function. prefrontal cortex (RPFC); rsfMRI: Resting-state fMRI; SPL: Superior parietal Proc Natl Acad Sci 107(10):4734–4739 lobule. 13. Smith SM et al (2009) Correspondence of the brain’s functional architec- ture during activation and rest. Proc Natl Acad Sci 106(31):13040–13045 Authors’ contributions 14. Smith SM et al (2013) Functional connectomics from resting-state fMRI. All authors read and approved the fnal manuscript. Trends Cogn Sci 17(12):666–682 15. Scheperjans F et al (2008) Probabilistic maps, morphometry, and vari- Funding ability of cytoarchitectonic areas in the human superior parietal cortex. This research work was funded by institutional fund projects under grant nos. Cereb Cortex 18(9):2141–2157 (IFPHI-327–142-2020) & (IFPHI-095-142-2020) & (IFPRC-004-142-2020) & (J: 16. Scheperjans F et al (2005) Subdivisions of human parietal area 5 revealed 67-142-1441). Therefore, authors gratefully acknowledge technical and fnan- by quantitative receptor autoradiography: a parietal region between cial support from the ministry of education and King Abdulaziz University, motor, somatosensory, and cingulate cortical areas. Neuroimage Deanship of Scientifc Research (DSR), Jeddah, Saudi Arabia. 25(3):975–992 17. Scheperjans F et al (2005) Transmitter receptors reveal segregation of Availability of data and materials cortical areas in the human superior parietal cortex: relations to visual All data used in the study are available in this website (https://​www.​nitrc.​org/​ and somatosensory regions. Neuroimage 28(2):362–379 proje​cts/​fcon_​1000/). 18. Dijk, K.R.a.V., et al (2010) Intrinsic functional connectivity as a tool for human connectomics: theory. Properties Optim 02138:297–321 19. O’Rawe JF, Ide JS, Leung H-C (2019) Model testing for distinctive func- Declarations tional connectivity gradients with resting-state fMRI data. Neuroimage 185:102–110 Ethics approval and consent to participate 20. Murphy K, Bodurka J, Bandettini PA (2007) How long to scan? The rela- The data sample were taken from the Cambridge-Buckner data sample tionship between fMRI temporal signal to noise ratio and necessary scan (open access) (https://​www.​nitrc.​org/​proje​cts/​fcon_​1000/). All procedures duration. Neuroimage 34(2):565–574 performed in the studies involving human participants were in accordance 21. Whitfeld-Gabrieli S, Nieto-Castanon A (2012) Conn: a functional con- with the ethical standards of the institutional and/or national research com- nectivity toolbox for correlated and anticorrelated brain networks. Brain mittee and with the 1964 Helsinki Declaration and its later amendments or connectivity 2(3):125–141 comparable ethical standards. 22. Eickhof SB et al (2005) A new SPM toolbox for combining probabilistic cytoarchitectonic maps and functional imaging data. Neuroimage Consent for publication 25(4):1325–1335 Informed consent was obtained from all individual participants included in 23. Scheperjans F et al (2008) Observer-independent cytoarchitectonic map- the study. The included participants data were taken from the Cambridge- ping of the human superior parietal cortex. Cereb Cortex 18(4):846–867 Buckner data sample, which is part of the 1000 functional connectomes 24. Jafri MJ et al (2008) A method for functional network connectivity among project. spatially independent resting-state components in . Neuro- image 39(4):1666–1681 Competing interests The authors declare that they have no confict of interest. Alahmadi Insights Imaging (2021) 12:47 Page 12 of 12

25. Benjamini Y, Hochberg Y (1995) Controlling the false discovery rate: a 43. Cui SZ et al (2000) Both sides of human cerebellum involved in practical and powerful approach to multiple testing. J Roy Stat Soc: Ser B preparation and execution of sequential movements. NeuroReport (Methodol) 57(1):289–300 11(17):3849–3853 26. Chumbley J et al (2010) Topological FDR for neuroimaging. Neuroimage 44. Imamizu H et al (2000) Human cerebellar activity refecting an acquired 49(4):3057–3064 internal model of a new tool. Nature 403(6766):192–195 27. Mars RB et al (2011) Difusion-weighted imaging tractography-based 45. Schlerf, J., et al., Big challenges from the little brain—imaging the cerebel- parcellation of the human parietal cortex and comparison with lum. Advanced brain neuroimaging topics in health and disease—meth- human and macaque resting-state functional connectivity. J Neurosci ods and applications, 2014. 31(11):4087–4100 46. Stoodley CJ (2012) The cerebellum and cognition: evidence from func- 28. Behrmann M, Geng JJ, Shomstein S (2004) Parietal cortex and attention. tional imaging studies. Cerebellum (London, England) 11(2):352–365 Curr Opin Neurobiol 14(2):212–217 47. Menon V, Uddin LQ (2010) Saliency, switching, attention and control: a 29. Corbetta M et al (1993) A PET study of visuospatial attention. J Neurosci network model of insula function. Brain Struct Funct 214(5–6):655–667 13(3):1202–1226 48. Strother SC (2006) Evaluating fMRI preprocessing pipelines. IEEE Eng Med 30. Shulman GL et al (2010) Right hemisphere dominance during spatial Biol Mag 25(2):27–41 selective attention and target detection occurs outside the dorsal fron- 49. Wang J et al (2005) To smooth or not to smooth? ROC analysis of perfu- toparietal network. J Neurosci 30(10):3640–3651 sion fMRI data. Magn Reson Imaging 23(1):75–81 31. Li G et al (2004) Cortical activations upon stimulation of the sensorimo- 50. Hagler DJ Jr, Saygin AP, Sereno MI (2006) Smoothing and cluster thresh- tor-implicated acupoints. Magn Reson Imaging 22(5):639–644 olding for cortical surface-based group analysis of fMRI data. Neuroimage 32. Premji A, Rai N, Nelson A (2011) Area 5 infuences excitability within the 33(4):1093–1103 primary motor cortex in humans. PLoS ONE 6:5 51. Wu CW et al (2011) Empirical evaluations of slice-timing, smoothing, and 33. Raichle ME (2015) The brain’s default mode network. Annu Rev Neurosci normalization efects in seed-based, resting-state functional magnetic 38:433–447 resonance imaging analyses. Brain connectivity 1(5):401–410 34. Greicius MD et al (2004) Default-mode network activity distinguishes 52. Alahmadi AA (2020) Efects of diferent smoothing on global and regional Alzheimer’s disease from healthy aging: evidence from functional MRI. resting functional connectivity. Neuroradiology 2020:1–11 Proc Natl Acad Sci 101(13):4637–4642 53. Molloy EK, Meyerand ME, Birn RM (2014) The infuence of spatial resolu- 35. Uddin LQ et al (2008) Network homogeneity reveals decreased integrity tion and smoothing on the detectability of resting-state and task fMRI. of default-mode network in ADHD. J Neurosci Methods 169(1):249–254 Neuroimage 86:221–230 36. Whitfeld-Gabrieli S, Ford JM (2012) Default mode network activity and 54. Filippi M et al (2013) The organization of intrinsic brain activity difers connectivity in psychopathology. Annu Rev Clin Psychol 8:49–76 between genders: A resting-state fMRI study in a large cohort of young 37. Uddin LQ et al (2009) Functional connectivity of default mode network healthy subjects. Hum Brain Mapp 34(6):1330–1343 components: correlation, anticorrelation, and causality. Hum Brain Mapp 55. Birn RM et al (2013) The efect of scan length on the reliability of resting- 30(2):625–637 state fMRI connectivity estimates. Neuroimage 83:550–558 38. Rosenberg-Lee M et al (2011) Functional dissociations between four 56. Noble S et al (2017) Infuences on the test–retest reliability of functional basic arithmetic operations in the human posterior parietal cortex: a connectivity MRI and its relationship with behavioral utility. Cereb Cortex cytoarchitectonic mapping study. Neuropsychologia 49(9):2592–2608 27(11):5415–5429 39. Chen G et al (2011) Negative functional connectivity and its depend- 57. Deco G, Corbetta M (2011) The dynamical balance of the brain at rest. ence on the shortest path length of positive network in the resting-state Neuroscientist 17(1):107–123 human brain. Brain Connectivity 1(3):195–206 58. Damoiseaux JS, Greicius MD (2009) Greater than the sum of its parts: a 40. Alahmadi AA et al (2016) Complex motor task associated with non-linear review of studies combining structural connectivity and resting-state BOLD responses in cerebro-cortical areas and cerebellum. Brain Struct functional connectivity. Brain Struct Funct 213(6):525–533 Funct 221(5):2443–2458 59. Friston KJ (2011) Functional and efective connectivity: a review. Brain 41. Alahmadi AA et al (2017) Cerebellar lobules and dentate nuclei mirror Connectivity 1(1):13–36 cortical force-related-BOLD responses: Beyond all (linear) expectations. Hum Brain Mapp 38(5):2566–2579 42. Alahmadi AA et al (2015) Diferential involvement of cortical and cerebel- Publisher’s Note lar areas using dominant and nondominant hands: An FMRI study. Hum Springer Nature remains neutral with regard to jurisdictional claims in pub- Brain Mapp 36(12):5079–5100 lished maps and institutional afliations.