Abstract
Objective:
Maldevelopment of the posterior corpus callosum is common in spina bifida myelomeningocele due to hydrocephalus-related hypoplasia and congenital partial hypogenesis. This study examined the relations of macro- and microstructural integrity of the interhemispheric temporal tract in spina bifida myelomeningocele and auditory interhemispheric transfer using consonant-vowel dichotic listening.
Method:
We collected T1-weighted and diffusion tensor imaging data from 46 individuals with SBM and 15 typically developing individuals. Probabilistic tractography was used to isolate the interhemispheric white matter connecting auditory processing regions in both hemispheres. Interhemispheric transfer was assessed with a dichotic listening task.
Results:
Although the typically developing group and the group with spina bifida myelomeningocele showed the normative right ear advantage, fewer participants showed a right ear advantage in the group with spina bifida myelomeningocele. The absence of the right ear advantage was largely in the subgroup with hypogenesis of the splenium or severe posterior hypoplasia. Sex, anterior commissure cross-sectional area, and number of shunt pathways visible on MRI predicted right ear superiority.
Conclusions:
Interhemispheric transfer is disrupted in individuals with spina bifida myelomeningocele and hypogenesis or severe hypoplasia of the posterior corpus callosum. Preservation of interhemispheric transfer is related to expected connections through the posterior corpus callosum and possibly compensatory pathways in the anterior commissure.
Keywords: spina bifida, dichotic listening tests, agenesis of corpus callosum, diffusion tensor imaging, white matter, neurodevelopmental disorders
Spina bifida myelomeningocele (SBM) is the most common and severe form of spina bifida (Copp et al., 2015). SBM is characterized by the formation of a lesion in which the spinal cord and meninges protrude through an opening in the spine (Anderson et al., 2001). This lesion occurs due to a failure of the neural tube to close properly during neuroembryogenesis, resulting in a cascade of malformations of the central nervous system (CNS) that affect physical, cognitive, and adaptive functioning (Dennis et al., 2006). At the level of the brain, SBM involves a primary deficit in neurulation and a secondary deficit due to the toxic effects of in utero CNS exposure to CSF. The spinal defect develops pockets of CSF that have a suction effect, leading to hindbrain herniation and the Chiari II malformation, the latter associated with hydrocephalus (Dewan & Wellons, 2019).
Although complete agenesis of the corpus callosum (CC) is rare in SBM, partial hypogenesis and hydrocephalus-related hypoplasia occur in over 90% of children with SBM (Hannay et al., 2000). Due to the timing of the formation of the spinal lesion during neuroembryogenesis, the posterior CC might be abnormal because of its later differentiation (Siffredi et al., 2013; Westerhausen et al., 2009). In addition, the CC can be damaged (thinned) by hydrocephalus secondary to the Chiari II malformation, which prior to the advent of prenatal surgery was virtually ubiquitous in SBM (Copp et al., 2015). Little is known about the connectivity of these posterior brain regions in individuals with SBM and how the macro- and microstructure of these interhemispheric white matter tracts relate to tasks that require interhemispheric transfer. In the absence of the formation of the posterior CC, interhemispheric white matter might be re-routed through other commissures (Hannay et al., 2009), but this topic remains underexplored.
On average, comparisons within the group show that individuals with SBM demonstrate relatively better performance on cognitive functions involving associative rule-based processes and weaknesses on tasks that require the assembling (construction) of knowledge (Dennis et al., 2006). For example, in language processing, children with SBM have strengths in lexical knowledge, but impaired higher-level language skills that require the extrapolation of meaning and the integration of complex ideas as evidenced by impaired idiom comprehension (Huber-Okrainec et al., 2005). In Huber-Okrainec and colleagues, the weakness in abstract idiom comprehension was related to the compromised integrity of the posterior corpus callosum, which hindered interhemispheric communication.
In a series of studies, Hannay and colleagues (Hannay, 2000, Hannay et al., 2008; 2009) examined the link between CC development and impaired interhemispheric transfer in SBM by evaluating performance on tasks that require interhemispheric transfer of information, such as dichotic listening (Hannay et al., 2008). On a consonant vowel (CV) dichotic listening task, right handers recall more consonants in the right than left ear (Satz, 1977). The right ear advantage (REA) varies with participant characteristics, with less asymmetry in non-righthanders, women, and those lower in SES (Bryden, 1988; Voyer, 2011). Using a CV task, Hannay et al. (2008) found that typically developing (TD) individuals and those with SBM and a normal or hypoplastic (thinned) splenium displayed the expected REA. However, individuals with hypogenesis (congenital underdevelopment) of the splenium failed to show the expected REA. The absence of the REA could be related to stronger ipsilateral connections, or the right hemisphere could be more involved in processing due to anomalous development. It is also possible that reorganization of white matter through alternate commissures may have preserved some interhemispheric transfer in people with partial hypogenesis.
A case study of two boys with partial dysgenesis of the CC found that an enlarged anterior commissure (AC) might be the source of re-routed interhemispheric auditory fibers (Fischer et al., 1992). However, Hannay et al. (2009), using radiological coding of the MRI in a sample of 193 children with SBM, found that alternate connections may not be responsible for compensated function. Enlargement of the AC was rare (3.1%). Probst bundles, which are callosal fibers that run longitudinally instead of crossing the CC, were also rare. Interhemispheric white matter pathways can be explored in greater detail with diffusion tensor imaging (DTI) than in these earlier studies.
In a previous structural imaging study not involving individuals from the studies by Hannay et al. (2008; 2009), Bradley et al. (2016) used DTI and probabilistic tractography to examine the macro- and microstructure of the CC in 76 individuals with SBM and 27 TD individuals. In people with SBM and partial hypogenesis or hypoplasia of the CC, superior temporal interhemispheric connections were passed (or projected) through alternate routes, including more anterior sections of the CC and the AC. More severe underdevelopment of the CC was associated with lower volume of the superior temporal interhemispheric tract (i.e., fewer interhemispheric fibers) and, based on DTI metrics, reduced microstructural integrity. Since function was not assessed, it is unclear if these altered connections were compensatory or maladaptive in SBM. To address this issue, we evaluated the relation between the structure of the posterior CC (normal, hypoplastic, partial hypogenesis) and performance on a CV dichotic listening task in SBM. Using DTI measures of macrostructure and microstructure of the superior temporal interhemispheric tract we also assessed whether these indices of white matter integrity were associated with performance on the dichotic listening task. We hypothesized:
Based on Hannay et al. (2008), both TD individuals and those with SBM will show a REA on the dichotic listening task. However, within the group with SBM, correct ear reports and the REA will vary depending on the status of the posterior CC (hypoplastic vs. severe hypogenesis) as well as other demographic (sex, SES) and clinical variables (lesion level, shunt) that may affect ear asymmetries. A reduced REA and altered left and right ear performance will be apparent in participants with SBM and hypogenesis or severe hypoplasia of the posterior CC.
Within the group with SBM, DTI indices of macrostructure (volume of the interhemispheric superior temporal tract and AC cross-sectional area) and microstructure (i.e. fractional anisotropy, FA; axial diffusivity, AD; and radial diffusivity, RD) of the interhemispheric temporal tract, will predict the number of correct ear responses.
Method
Participants
The sample included 61 children and adults between 8 and 43 years old, 46 with SBM and 15 TD individuals. All participants were a subset of those who also had received neuroimaging and completed the dichotic listening task in the neuroimaging investigation by Bradley et al. (2016), which should be consulted for a more detailed exposition of the neuroimaging methods and results. Participants with SBM were recruited from the spina bifida clinics at Texas Children’s Hospital and the Shriner’s Hospital for Children in Houston. The comparison group of TD individuals represented volunteers recruited through advertisement from the community. The protocol was approved by Institutional Review Boards at The University of Houston and The University of Texas Health Science Center-Houston. Adolescents and adults 13 and older gave written informed consent. Those under 13 assented to the study. Parents of all participants under 18 gave their written informed consent for participation.
Exclusionary criteria included any of the following: other genetic, neurodevelopmental, or severe psychiatric disorders (e.g., autism spectrum disorder), and an uncontrolled seizure disorder. All participants had a verbal or visual reasoning composite score of at least 70 on the Stanford-Binet Intelligence Test-IV (Thorndike et al., 1986) to ensure the ability to follow directions and perform the DLT. The TD group was solely right-handed as assessed by hand preference on the Beery Test of Visual-Motor Integration (Beery, 1982) in order to control for language lateralization (Bryden, 1988; Hannay et al., 2008). The group with SBM was not restricted by handedness because early brain injury often results in non-righthandedness and SBM has a higher rate of non-righthandedness than the general population (38% in Fletcher et al., 2005). A measure of hand preference for drawing was used because we wanted to make a binary decision of right versus non-righthandedness for reporting clinical characteristics; handedness was not quantitatively analyzed because a performance measure of handedness was not obtained and the binary variable underestimates the proportion of individuals with mixed handedness. SES was assessed with the Hollingshead 4-Factor Scale (Hollingshead, 1975), a composite of parent education and occupational status. Participants were excluded for subtle hearing difficulties with ≥ 20 db difference between ears or thresholds ≥ 60 db in each ear at each frequency for hearing pure tones monaurally (500, 1000, 2000, 4000 Hz) presented with a Beltone Portable 100 Series Model Audiometer (Beltone Electronics, Glenview, IL). One participant with SBM was dropped because of extremely low performance on the DLT (i.e., outlier).
Transparency and Openness
We reported how we determined our sample size, all data exclusions, all manipulations, and all relevant measures in the study, and we followed JARS. All data, analysis code, and research materials will be available. Data were analyzed using SPSS Software, version 22. This study’s design and its analysis were not pre-registered.
Procedures
Dichotic Listening
Stimuli were presented using a TASCAM 202 MKII cassette deck with an Optimus SA-155 stereo amplifier (Hannay et al., 2008). Participants wore Sony MDR-7506 professional stereo headphones that were calibrated to an output level of 81 dB to listen to the presented CV syllables. The participants were told to listen to two syllables at the same time, one in each ear, and to report the clearer syllable first, and then the other syllable (Hannay et al., 2008). The test administrator also wore a comparable set of headphones and recorded responses.
Monotic Listening.
A monotic listening task was given solely to ensure that participants could distinguish among the 6 syllables used in the DLT (Hannay et al., 2008). Single CV pairs (i.e., /ba/, /da/, /ga/, /ka/, /pa/, and /ta/) were presented one at a time. Each item was randomly repeated three times for each ear (18 trials to the left ear and 18 trials to the right ear), for a total of 36 monotic trials (Hannay et al., 2008).
Dichotic Listening.
Two different CV syllables from the six syllables used for the monotic task were presented simultaneously to the right and left ears for 36 trials. The headphones were then reversed for an additional 36 trials to control for differences in headphone output (Hannay et al., 2008). The number of correct first responses for stimuli presented to the right ear and left ear was assessed because first (clearest) responses are less subject to guessing (Hannay et al., 2008). Laterality indices were not used in analyzing the dichotic listening data, instead favoring the evaluation of individual ear contributions and a simple difference between ears (i.e. right – left) (Springer, 1986; Hannay et al., 2008).
MRI Acquisition
MR images were acquired using a 3T Philips Intera scanner with SENSE (Sensitivity Encoding) technology. High-resolution T1-weighted anatomical images were acquired in the coronal plane using a 3D turbo fast echo sequence with the following parameters: voxel dimensions = .94 x .94, slice thickness = 1.5 mm, TR = 6.50–6.70 ms, TE = 3.04–3.14 ms, flip angle = 8°, DFOV = 240 mm, matrix = 256 × 256. DTI images were acquired in the axial plane using a spin-echo diffusion sensitized echo-planar imaging sequence. Diffusion sensitizing gradients were applied in 21 directions (weighting: b = 1000 s/mm2) with one reference image (b = 0 s/mm2) and the following parameters: voxel dimensions = .94 x .94, slice thickness = 3 mm, TR=6500 ms, TE=65 ms, flip angle = 90°, DFOV = 240 mm, matrix = 256 × 256.
MRI Data Analysis
Radiological Coding of T1-Weighted Images.
Each participant’s CC (rostrum, genu, body, and splenium) was qualitatively classified as present, absent, or hypoplastic by a radiologist blind to the status of participants’ status using T1- weighted MRI data. Three subgroups were formed from this procedure (Bradley et al., 2016): the partial hypogenesis subgroup constituted the most severe, with the splenium absent or severely shortened/thinned; the hypoplastic subgroup included participants where the splenium was present but mildly or moderately thinned; and the intact subgroup included participants with all portions of the CC present and normal appearing.
DTI Probabilistic Tractography.
T1- weighted data underwent cortical parcellation and transformation into diffusion space in order to extract seed masks for use in the probabilistic tractography analyses (Bradley et al., 2016). Two seed masks were chosen for probabilistic tractography, one in the left and another in the right posterior temporal lobe where Heschl’s gyrus and the superior temporal lobe meet. Intrarater reliability of the seed masks was assessed through calculation of the Dice similarity coefficient on 10% of the left and right masks selected at random. High intrarater reliability was shown with a Dice similarity coefficient of 0.93 (SD = 0.027), with a range of 0.86 to 0.96 (Williams et al., 2013).
DTI data were preprocessed for probabilistic tractography using FMRIB’s Software Library (FSL) version 5.0.1 (Jenkinson et al., 2012; Smith et al., 2004; Woolrich et al., 2009) and FSL’s PROBTRACKX. Two fiber tracts were created, one seeded from the left posterior temporal lobe with the right posterior temporal lobe as a waypoint, and another from the right posterior temporal lobe with the left posterior temporal lobe as a waypoint. An exclusion mask of the midbrain was included in order to prevent tracking the ascending and descending auditory fibers that connect the left- and right-hemisphere temporal lobes. The two tracks were combined through FSL terminal commands that multiplied the binarized tracts together to only keep voxels that were shared. All DTI metrics (FA, AD, RD, and total tract volume) were extracted from this combined tract in order to ensure the most stringent criteria for selecting voxels that exist along the white matter path were met (Javad et al., 2014; Jones, 2011). Measures of FA, AD, and RD were extracted from the combined tract because these metrics are thought to measure microstructural changes in white matter. Common interpretations suggest that FA is an overall measure of microstructural integrity of brain structures, but not necessarily specific to the type of pathological changes that can occur such as demyelination or axonal degeneration (Alexander et al., 2008). However, evaluation of parallel (AD) and perpendicular diffusion (RD) may further allow for more specific conclusions to be drawn about underlying microstructural changes. For example, typical white matter maturation is associated with increases in FA and AD, but reductions in RD (Alexander et al., 2008). Axonal degeneration is characterized by an opposite pattern of decreased FA and AD, and increased RD (Alexander et al., 2008). Lastly, demyelination is characterized by decreased FA, little to no change in AD, and increased RD (Alexander et al., 2008).
Based on where the combined interhemispheric temporal tract crossed the midline, patterns of interhemispheric connection were numerically coded 1–7 for analyses; detailed descriptions of these tractography methods, as well as visual representation of these patterns of connectivity can be seen in Bradley et al. (2016). The coded patterns of connections included the following: 1 = the tract crossed through the posterior CC only; 2 = the tract crossed through the posterior CC and AC; 3 = the tract crossed through the AC only; 4 = the tract crossed through the anterior and posterior CC; 5 = the tract crossed through the anterior CC, posterior CC, and AC; 6 = the tract crossed through the anterior CC and AC; 7 = the tract crossed through the small CC remnant and AC.
Calculation of CC Volume and AC Cross-Sectional Area.
A quantitative measure of total CC volume was computed to assess the degree of CC underdevelopment. Deterministic tractography was performed with the fiber assignment by continuous tracking (FACT) algorithm in TrackVis (http://trackvis.org); the complete procedure can be found in Bradley et al. (2016). This procedure involved seeding all white matter in the brain and only keeping those fibers that crossed through a manually drawn ROI of the CC as in previous investigations (Wahl et al., 2009). Additionally, using the midsagittal slice of the brain, a manual ROI was drawn over the AC. The voxel dimensions (mm) in the coronal plane and axial plane within only the masked area were multiplied together to get an estimate of cross-sectional area of the AC (in mm2).
Results
Demographics
Demographic information was presented in Table 1. Participants in the TD (n = 15) and SBM (n = 46) groups did not differ significantly in age, sex, ethnicity, or SES (p > .05). As expected, the TD group scored higher on both verbal, t(59) = 2.43, p = 0.018, and visual, t(59) = 2.54, p = 0.014, Stanford-Binet composites than those with SBM.
Table 1.
Demographics by Group
| TD | SBM | |
|---|---|---|
|
| ||
| N | 15 | 46 |
| Age in years: M(SD) | 19.99 (10.29) | 16.24 (6.35) |
| Sex: N (% men) | 6 (40.00) | 28 (60.90) |
| Socioeconomic status: M(SD)1 | 37.50 (10.37) | 31.20 (12.48) |
| Handedness: N (% Right) | 15 (100) | 37 (80.40) |
| Ethnicity: N (%)1 | ||
| % Hispanic | 9 (60.00) | 23 (50.00) |
| % Non-Hispanic | 6 (40.00) | 22 (47.80) |
| Stanford Binet IQ | ||
| Verbal Reasoning: M(SD) | 96.67 (13.43) | 86.13 (14.93)* |
| Visual Reasoning: M(SD) | 102.13 (14.59) | 92.44 (12.26)* |
Note:
p < 0.05; TD = Typically developing; SBM = Spina bifida myelomeningocele;
Missing data on one participant
Clinical Characteristics
In the group with SBM, the majority of participants had lower-level (lumbar or sacral) spinal lesions (78%) and presented with a Type II Chiari malformation (89%). Additionally, 73% presented with hypoplastic CCs and 24% showed partial hypogenesis/severe hypoplasia. Few participants had past or current seizure disorders (9%). Only 2% of participants showed no shunting. Of those with shunting, the majority had only one shunt pathway visible on MR images (74%).
Monotic Listening
One sample t-tests showed that CV syllable identification was well above chance (i.e., identifying 3 out of 6 CV syllables) for all participants (left ear: t(60) = 36.02, p < .001; right ear: t(60) = 34.41, p < .001). Both the TD group (n = 15) and group with SBM (n = 46) could discriminate among the stimuli used in the dichotic listening test. Mann Whitney U-Tests revealed that median left ear scores [U = 264, z = −1.374, p = 0.17], median right ear scores [U = 306.5, z = −0.653, p = 0.51], and the difference between participants’ right and left ear scores [right-left; U = 423.5, z = 1.343, p = 0.18] were not significantly different between the TD and SBM groups. Table 2A displays monotic listening data.
Table 2.
| A) Monotic Listening Performance in TD and SBM | |||
|---|---|---|---|
|
| |||
| TD (n = 15) | SBM (n = 45) | Mann Whitney U-Test | |
|
| |||
| Left ear mean (SD) [Range] | 14.67 (2.29) [12–18] | 13.63 (2.35) [8–18] | p = 0.17 |
| Right ear mean (SD) [Range] | 13.93 (2.25) [11–18] | 13.33 (2.42) [8–18] | p = 0.51 |
| B) Dichotic Listening Performance in TD and SBM | ||
|---|---|---|
|
| ||
| TD (n = 15) | SBM (n = 45) | |
|
| ||
| Left Ear Mean (SD) | 26.87 (4.52) | 24.31 (8.50) |
| Right Ear Mean (SD) | 33.00 (5.64) | 32.82 (9.49) |
| Total Correct Mean (SD) | 59.87 (5.07) | 57.13 (7.37) |
| Contrast | R > L* | R > L* |
Note: R = right; L = left; TD= typically developing; SBM = spina bifida myelomeningocele
p < .05
Hypothesis 1: Dichotic Listening Performance: SBM vs. TD
One sample t-tests evaluated whether participants in each group scored above chance performance for each ear on the dichotic listening task. If there was no ear preference, each ear could have a total maximum of 36 correct first responses on the 72 trials. Therefore, with 6 choices of syllables (i.e. /ba/, /da/, /ga/, /ka/, /pa/, and /ta/) on each trial, chance level would be 6 correct responses (i.e. 36 divided by 6). Both the TD group (left ear responses: t(14) = 17.89, p < .0005; right ear responses: t(14) = 18.53, p < .0005) and the group with SBM (left ear responses: t(44) = 14.45, p < .0005; right ear responses: t(14) = 18.96, p < .0005) scored significantly above chance on both the number of correct left and right ear responses.
Unlike the TD group in which monotic performance was not related to dichotic performance, correct right ear responses on the monotic task were moderately positively correlated with both the right ear score on the dichotic task in the group with SBM, r = .37, p = .012, and the dichotic difference score (i.e. right ear – left ear), r = .318, p = .033. The standard deviation in the group with SBM was twice the standard deviation in the TD group, which may contribute to the correlation in the group with SBM because of increased range of performance. The left and right ear monotic scores were strongly positively correlated, r = .67, p < .0005.
Overall Group (TD, SBM) Performance
Table 2b displays mean left and right ear reports for both groups, as well as a measure of overall performance on the DLT (correct left ear responses + correct right ear responses). A univariate ANCOVA, controlling for age, found that the TD group (n = 15) and the group with SBM (n = 45) did not differ significantly on overall performance on the DLT, F(1,57) = 1.12, p = .334, η2 = .04. The interaction of age in the model was not significant, and when trimmed, results were unchanged, with no significant difference in overall DLT performance between the TD and SBM groups, F(1,58) = 1.77, p = .188, partial eta2 = .03. The TD group was able to discriminate a mean of 59.87 (SD = 5.07) syllables correctly, while the SBM group scored almost as well (mean = 57.13, SD = 7.37).
Within-subjects ANOVAs, with ear as a repeated factor, compared performance on the DLT in the TD group and the group with SBM separately in order to determine if either group showed a REA. In the TD group, the normative right ear advantage (REA) was significant, F(1,14) = 7.16, p = .018, η2 = .338. In the group with SBM, there was also a significant REA, F(1,44) = 12.06, p = .001, η2= .215.
Subgroups with SBM.
Mixed model ANCOVAs were performed to evaluate interactions of correct left and right ear reports and various demographic (sex, SES, age), structural (e.g., radiological coding of the posterior CC as normal, hypoplastic, hypogenetic), and clinical features (lesion level, shunt-related variables) within the group with SBM. Preliminary mixed ANCOVA models showed no statistically significant (p > .05) effects of age, SES, or sex and either left or right ear effects, so these variables were trimmed to preserve degrees of freedom. Number of visible shunt pathways, number of shunt revisions, lesion level, and whether the shunt transversed the CC showed no significant interactions with ear or significant main effects (p > .05).
Ear by Structure of the Posterior CC.
The participant with SBM and a normal posterior CC (n=1) was left out of these group comparisons. Dichotic listening performance for the subgroups with SBM and either a hypoplastic CC ( n = 33) or more severe partial hypogenesis of the CC (n = 11) is presented in Figure 1. The ear by status of the splenium (i.e. hypoplastic, partial hypogenesis) ANOVA revealed a significant interaction, F(1,42) = 5.70, p = .022, η2 = .12. Simple effects Bonferroni follow-up tests (.05/4, critical alpha = .0125) revealed that the hypoplastic subgroup had significantly more correct right ear responses than the severely hypoplastic/partial hypogenesis subgroup. The hypoplastic CC subgroup scored a mean of 35.03 (SD = 9.05) correct right ear reports, while the severely hypoplastic/partial hypogenesis CC subgroup scored a mean of 26.64 (SD = 8.59) right ear reports; right ear reports were decreased by a mean of 8.39 correct responses in the group with more severe hypoplasia/hypogenesis. There were no differences between subgroups of SBM on the number of correct left ear reports (p > .05).
Figure 1.
Dichotic Listening Performance for Subgroups With SBM and Either a Hypoplastic or More Severe Partial Hypogenesis of the CCNote. SBM = spina bifida myelomeningocele; CC = corpus callosum. * p < .05.
Additional planned Bonferroni comparisons (.05/4, critical alpha = .0125) showed that within the hypoplastic subgroup, the correct number of right ear reports (M = 35.03, SD = 9.05) exceeded left ear reports (M = 23.18, SD = 8.50). Individuals with SBM and a hypoplastic posterior CC showed the expected REA, whereas those with hypogenesis did not show the expected response pattern (p = .791). There was little difference between left (M = 27.91, SD = 8.25) and right ear reports (M = 26.64, SD = 8.59) in the group with more severe hypogenesis/hypoplasia; this group showed a small, non-significant left ear advantage.
Hypothesis 2: Quantitative DTI Indices and Dichotic Performance in SBM
A multiple regression model evaluated predictors of right ear superiority (i.e., the right ear advantage) in dichotic performance. A number of covariates were considered, including age, sex, AC cross-sectional area, number of visible shunt pathways and revisions, FA, AD, RD, volume of the interhemispheric temporal tract, and total CC volume. All regression results are presented in Table 3.
Table 3.
| A) Significant predictors of dichotic right ear superiority (REA) in SBM | ||||
|---|---|---|---|---|
|
| ||||
| B | SEB | Beta | p | |
|
| ||||
| Sex | −9.827 | 4.576 | −.296 | .039* |
| AC size | 1.264 | .367 | .482 | .002* |
| Number of Shunts | −10.386 | 3.513 | −.447 | .006* |
|
| ||||
| B) Non-significant predictors of dichotic right ear superiority (REA) in SBM | ||||
|
| ||||
| B | SEB | Beta | p | |
|
| ||||
| Age | .102 | .378 | .040 | .790 |
| FA | 72.365 | 251.118 | .362 | .775 |
| AD | −3.512 | 87.583 | −.033 | .968 |
| RD | 52.773 | 156.045 | .565 | .737 |
| Volume of temporal tract | .001 | .003 | .052 | .746 |
| Total CC volume | 2.098 | 1.552 | .218 | .185 |
Note: SBM = spina bifida myelomeningocele; AC = anterior commissure; AD = axial diffusivity; CC = corpus callosum; RD = radial diffusivity; REA = right ear advantage
p < .05
The REA was calculated as a difference score (i.e. correct right ear responses – correct left ear responses). The multiple regression model that included age, sex, AC cross-sectional area, number of visible shunt pathways, FA, AD, RD, volume of the interhemispheric temporal tract, and total CC volume was significant, F (9, 44) = 3.09, p = .008, R2 = .443. Within this model, sex, number of visible shunt pathways, and AC cross-sectional area were significant predictors of the REA. Being a man (Beta = −.296), having fewer visible shunt pathways (Beta = −.447), and a larger AC cross-sectional area (Beta = .482) were all associated with a larger REA. Age, FA, AD, RD, volume of the temporal tract, and CC volume were not significant predictors of the REA.
Follow Up Analyses
In the spirit of the case studies that historically drive the compensatory question, the Appendix presents the pattern of interhemispheric temporal lobe connections for 59 participants in relation to ear superiority (i.e. REA) on the dichotic listening task, age, and handedness. The outlier and the individual with SBM and a normal CC were not included in this summary.
TD Group.
In the TD group, the Appendix (A), 80% showed a REA, while 20% showed a LEA, which is consistent with normative expectations (Satz, 1977). All participants in the TD group were right-handed, so the LEA was not due to non-righthandedness. Additionally, all participants showed connections between the posterior temporal lobes through only the posterior CC. These results suggest that even in TD there is variability in dichotic performance such that not all people show the expected REA despite displaying the typical pattern of connection through the posterior CC.
Hypoplastic subgroup.
Within the group with SBM, the majority of people with hypoplasia (Appendix, B), showed a REA (73%) as expected. When tract location was examined in relation to ear superiority, it was found that both those with a REA and those with a LEA showed similar proportions of people with connections through the posterior CC (over 80% with either a REA or LEA). This preservation in posterior CC connectivity may explain why the majority of the hypoplastic subgroup shows a REA on the dichotic task. Furthermore, we found no relation between AC connectivity and performance, with equal numbers with a LEA and connections through the AC (n=3) versus a REA and connections through the AC (n = 3). Compensatory connectivity because of an enlarged AC is infrequent in SBM (3.1% in Hannay et al., 2009). However, the regression analyses showed that AC volume was related to the REA, suggesting that larger ACs might be compensatory.
Hypogenesis Subgroup.
In the group with SBM and more severe partial hypogenesis/hypoplasia of the CC (Appendix, C), there was a more even split in ear superiority, with only 55% demonstrating a REA. However, of those who did show a REA, 83% had connections through the most posterior part of the CC, while only 60% of individuals who showed a LEA had connections in the posterior CC.
Discussion
Few studies of neurodevelopmental disorders examine, much less demonstrate, associations of quantitative assessments of brain volume and integrity with behavioral performance. In this study, while the group with SBM showed an overall REA, subgroups with SBM and severe malformation of the posterior CC were less likely to show a REA, supporting our first hypothesis that more severe malformation of the posterior CC would adversely affect dichotic performance and replicating Hannay et al. (2008). Additionally, sex, AC cross-sectional area, and number of visible shunt pathways were associated with ear superiority on the DLT. Being a woman, smaller AC size, and more visible shunt pathways, a proxy for shunt revisions, were associated with reduced right ear superiority on the DLT.
Dichotic Listening Performance (Hypothesis 1)
The typical response for people with left-lateralized language on dichotic-listening paradigms is a REA, represented by reporting a greater number of stimuli from the right ear than the left (Kimura, 1961). According to Kimura’s structural theory, the organization of the auditory pathways explains this phenomenon because there are a greater number and stronger connections between a stimulated ear and the contralateral auditory cortex (Rosenzweig, 1951); therefore, if language is left lateralized, stimuli from the right ear will be reported more frequently because they have a more direct connection to left-lateralized auditory and language processing regions that does not require transfer through the CC.
Consistent with Hannay et al. (2008), we found that the REA is preserved in people with SBM and CC hypoplasia, which is a mechanical process that thins the CC because of hydrocephalus. However, individuals with congenital partial hypogenesis or severe hypoplasia of the posterior CC were less likely to show a REA.
Relation of Structure and Function (Hypothesis 2)
Cross-sectional area of the AC was the only macrostructural variable that significantly predicted right ear superiority on the dichotic listening task. Volume of the interhemispheric superior temporal tract and total CC volume did not show associations with dichotic performance. A larger AC area was associated with a larger REA. Although Hannay et al. (2009) found that an enlarged AC occurred in only 3.1% of individuals with SBM on qualitative assessment, the larger area on quantitative assessment is consistent with a compensatory function.
In terms of microstructure of the interhemispheric superior temporal tract, none of our hypothesized indices of integrity, including FA, AD, or RD, predicted performance on the dichotic listening task in SBM. From our results, it seems that markers of more diffuse disruption in brain development and damage, such as status of the posterior CC and number of visible shunt pathways, were more strongly associated with dichotic performance than microstructural indices of integrity of the superior temporal tract that is directly involved in the interhemispheric transfer of auditory information on the DLT. While unexpected, these results suggest that preservation of posterior CC connectivity is a primary mechanism for preserved interhemispheric transfer of auditory signals.
The relation with visible shunt pathways is interesting. Individuals with multiple shunt revisions tend to have more severe impairment of white matter (Williams et al., 2013). Counting shunt revisions retrospectively has weak association with cognitive skills (Arrington et al., 2016), so counting pathways from MRI might be more accurate than retrospective coding from medical history. The impairment of the CC in SBM is part of a more general pattern of disrupted white matter, which implies that the relations observed in this study may not be completely specific to the CC. On average, men have a slightly stronger REA than women on DLT (Voyer, 2011). This observation may account for the association of sex and DLT performance in this study.
Patterns of Connectivity and Dichotic Performance
As the Appendix shows, individual cases vary in patterns of callosal maldevelopment and connectivity, so generalizations from group data should be made with caution. Despite the relation of AC cross-sectional area and ear superiority on the DLT, both the hypoplastic and hypogenesis subgroups with SBM showed similar connections through the AC. However, there were no observed differences in the proportion of people that showed AC connection with either a REA or a LEA, so the larger area does not in itself suggest that the AC is compensatory. A more obvious mechanism is connectivity through the posterior CC, which is more preserved in the hypoplastic subgroup. Larger AC size is more likely in the less severe hypoplastic individuals, in which preservation of posterior CC connection is also more likely. However, whether these connections through the AC are truly compensatory is still unclear given that larger AC area did predict a REA on the DLT.
Contraints on Generality
There are technical limitations inherent to the DTI analyses. The confidence assigned to connections in probabilistic tractography diminishes with increasing distance from the starting seed point (Javad et al., 2014; Jones, 2011). This is why two tracts were created for each interhemispheric connection of interest, one seeded from the left temporal lobe and one from the right. By combining these reverse tracts and only keeping voxels that exist in both directions, this confound was minimized and we could be surer of the actual existence of the white matter tract of interest (Javad et al., 2014).
Not all white matter fibers connecting ROIs can be identified due to limitations in spatial resolution and sensitivity of DTI. The interhemispheric tracts cross many association tracts. This “crossing-fibers” problem is a known limitation of tractography. The DTI acquisition methods were state of the art when data collection began in 2006. Slice thickness of the DTI acquisition was 3mm and the voxels were not isotropic. Only 21 diffusion directions were applied, which further limited the probabilistic methodology. Inclusion of the tapetum in the hand drawn seed ROI helped to track the path of interest to mitigate this problem. In a study evaluating ascending and interhemispheric auditory pathways in healthy adults, probabilistic DTI tractography was only able to identify tracks connecting the left and right sound processing regions in the temporal lobes in 86% of hemispheres (Javad et al., 2014). In the current study, the tractography procedure worked to track the interhemispheric temporal connections in 89% of participants. This rate is in line with previous research (Javad et al., 2014), suggesting that the probabilistic method employed in the current study worked within the expectations of published literature despite the limitations in both the acquisition and analysis of the data.
There are also limitations to the interpretation of DTI metrics of integrity, specifically AD and RD. In clinical populations with severe neural pathology, the primary, secondary, and tertiary eigenvectors that define diffusion along the white matter tract may not actually align with actual tissue organization (Wheeler-Kingshott & Cercignani, 2009). This means that AD and RD values are not necessarily accurate predictors of parallel and perpendicular diffusion respectively, which limits the conclusions that can be drawn from these metrics in general.
Lastly, the small sample of people with SBM and severe partial hypogenesis of the CC reduced the ability to examine more complex statistical models, particularly in relation to individual differences in dichotic performance. Relations were not found between microstructural indices of integrity and dichotic performance. It is possible that with a larger sample size, more complicated relations between these variables could be examined, especially given that SBM is a heterogeneous disorder. Many statistical tests were completed and we did not always adjust the alpha threshold to control possible Type 1 errors for the subgroup analyses. SBM is a rare disease and we felt it more important to avoid missing potentially important relations because of a conservative approach to hypothesis testing. Effect sizes were reported for all statistical tests meeting the critical level of alpha.
Conclusions
These findings have implications beyond just one etiology or one type of disruption in neurodevelopment. Connections through alternative commissures such as the AC were found in people with both partial hypogenesis and hypoplasia of the CC, and our analyses do suggest some relation between AC size and ear superiority on the DLT. Whether these connections are truly compensatory is still unclear given the association between AC size and preservation of posterior CC connectivity in less severe cases of CC hypoplasia. Given that there are many different causes of both CC hypoplasia and hypogenesis that don’t involve SBM (Anderson et al., 2001), the issue of whether alternate routes of interhemispheric transfer develop to compensate for early developmental disruption in brain development remains underexplored in other neurodevelopmental disorders. SBM involves many CNS abnormalities, which may contribute to the maladaptive function of these aberrant connections. In other disorders, such as congenital dysgenesis where the CC doesn’t form but other brain structures are relatively intact, plastic connections through other commissures might serve more compensatory functions. It is possible that different patterns of disruption in neurodevelopment may lead to plastic re-routing of white matter connections, and these connections could be either compensatory or maladaptive.
Future studies should specifically examine how disruption in neurodevelopment relates to neuroplasticity and cognitive function in a broader context. Newer and more advanced DTI methods allow the tracking of pathways with higher spatial resolution that are more sensitive to some of the limitations of tractography. These methodologies will further push the boundaries on examining the relations between structure and function in neurodevelopmental disorders.
Key points.
Question:
What is the relation of the corpus callosum malformation and auditory interhemispheric transfer in spina bifida myelomeningocele
Findings:
Interhemispheric transfer is preserved in spina bifida if the corpus callosum is thinned, but not if it is congenitally malformed. Preservation depends on the integrity of callosal fibers in the posterior part of the corpus callosum and the anterior commisure.
Importance:
Interhemispheric transfer is primarily determined by preserved callosal tracts, with some evidence for compensatory functions in the anterior commissure.
Next steps:
The research should be expanded to other neurodevelopmental disorders with malformation of the corpus callosum.
Acknowledgments
The Eunice Kennedy Shriver National Institute of Child Health and Human Development (grant P01-HD35946–06 to J.M.F.) supported this work. The content of this manuscript is solely the responsibility of the authors and does not necessarily represent the official views of the Eunice Kennedy Shriver National Institute of Child Health and Human Development or the National Institutes of Health. Maureen Dennis passed away on July 14, 2014, but she helped to design the study and interpret the results. A presentation based on this paper was awarded the Guthkelch Award for best student presentation to Bradley by the Society for Research in Hydrocephalus and Spina Bifida, Sterling, U.K., 2016
Appendix
Participant listing of the difference in right and left ear dichotic performance, ear superiority on the dichotic task, age, handedness, and the location (marked by an “X”) where the auditory interhemispheric tract crossed hemispheres
| A) Interhemispheric tract location and dichotic performance for the TD group | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| TD (n = 15) | Interhemispheric Tract Location | ||||||||||
|
| |||||||||||
| n | DL Difference (R-L) | Ear Superiority | Age | Hand | Lesion Level | Post CC | Post CC & AC | AC Only | Post & Ant CC | Post & Ant CC & AC | Ant CC & AC |
|
| |||||||||||
| 1 | −13 | 29.27 | R | None | X | ||||||
| 1 | −9 | 20 % Left | 32.16 | R | None | X | |||||
| 1 | −1 | 8.29 | R | None | X | ||||||
|
| |||||||||||
| 1 | 2 | 10.16 | R | None | X | ||||||
| 1 | 2 | 30.98 | R | None | X | ||||||
| 1 | 5 | 10.04 | R | None | X | ||||||
| 1 | 7 | 15.15 | R | None | X | ||||||
| 1 | 8 | 23.00 | R | None | X | ||||||
| 1 | 9 | 80 % Right | 10.78 | R | None | X | |||||
| 1 | 10 | 15.69 | R | None | X | ||||||
| 1 | 10 | 36.11 | R | None | X | ||||||
| 1 | 11 | 35.64 | R | None | X | ||||||
| 1 | 16 | 10.17 | R | None | X | ||||||
| 1 | 17 | 12.40 | R | None | X | ||||||
| 1 | 18 | 19.94 | R | None | X | ||||||
|
| |||||||||||
| B) Interhemispheric tract location and dichotic performance in the group with SBM and callosal hypoplasia | |||||||||||
| SBM Hypoplastic Posterior CC (n = 33) | Interhemispheric Tract Location | ||||||||||
|
| |||||||||||
| n | DL Difference (R-L) | Ear Superiority | Age | Hand | Lesion Level | Post CC | Post CC & AC | AC Only | Post & Ant CC | Post & Ant CC & AC | Ant CC & AC |
|
| |||||||||||
| 1 | −16 | 18.03 | R | Lower | X | ||||||
| 1 | −13 | 20.05 | R | Lower | X | ||||||
| 1 | −12 | 12.44 | R | Lower | X | ||||||
| 1 | −12 | 21.20% | 18.36 | R | Lower | X | |||||
| 1 | −8 | Left | 12.34 | NR | Upper | X | |||||
| 1 | −7 | 17.79 | R | Upper | X | ||||||
| 1 | −1 | 9.48 | R | Lower | X | ||||||
|
| |||||||||||
| 1 | 0 | 6.06% | 11.09 | R | Lower | X | |||||
| 1 | 0 | None | 43.37 | NR | Lower | X | |||||
|
| |||||||||||
| 1 | 2 | 9.62 | NR | Upper | X | ||||||
| 1 | 3 | 10.47 | NR | Lower | X | ||||||
| 1 | 4 | 9.79 | NR | Lower | X | ||||||
| 1 | 8 | 11.92 | R | Lower | X | ||||||
| 1 | 8 | 15.07 | NR | Lower | X | ||||||
| 1 | 9 | 17.51 | R | Lower | X | ||||||
| 1 | 9 | 18.48 | R | Lower | X | ||||||
| 1 | 11 | 14.26 | R | Upper | X | ||||||
| 1 | 11 | 16.55 | R | Lower | X | ||||||
| 1 | 11 | 25.18 | R | Upper | X | ||||||
| 1 | 17 | 19.99 | R | Upper | X | ||||||
| 1 | 20 | 72.70% | 14.21 | R | Lower | X | |||||
| 1 | 20 | Right | 23.45 | R | Lower | X | |||||
| 1 | 21 | 12.78 | R | Lower | X | ||||||
| 1 | 23 | 18.96 | R | Lower | X | ||||||
| 1 | 24 | 16.96 | R | Lower | X | ||||||
| 1 | 25 | 25.39 | R | Lower | X | ||||||
| 1 | 27 | 20.34 | R | Lower | X | ||||||
| 1 | 28 | 9.40 | R | Upper | X | ||||||
| 1 | 28 | 22.72 | R | Lower | X | ||||||
| 1 | 29 | 10.89 | R | Lower | X | ||||||
| 1 | 34 | 16.18 | R | Lower | X | ||||||
| 1 | 41 | 16.04 | R | Lower | X | ||||||
| 1 | 47 | 9.50 | R | Lower | X | ||||||
|
| |||||||||||
| C) Interhemispheric tract location and dichotic performance in the group with SBM and partial hypogenesis of the CC | |||||||||||
| SBM Partial Hypogenesis of the Posterior CC (n = 11) | Interhemispheric Tract Location | ||||||||||
|
| |||||||||||
| n | DL Difference (R-L) | Ear Superiority | Age | Hand | Lesion Level | Post CC | Post CC & AC | AC Only | Post & Ant CC | Post & Ant CC & AC | Ant CC & AC |
|
| |||||||||||
| 1 | −24 | 11.21 | NR | Lower | X | ||||||
| 1 | −21 | 45.5% | 24.00 | R | Lower | X | |||||
| 1 | −13 | Left | 12.52 | R | Lower | X | |||||
| 1 | −10 | 12.88 | R | Lower | X | ||||||
| 1 | −1 | 11.25 | R | Upper | X | ||||||
|
| |||||||||||
| 1 | 1 | 14.08 | R | Lower | X | ||||||
| 1 | 3 | 10.43 | NR | Lower | X | ||||||
| 1 | 5 | 54.5% | 14.45 | R | Lower | X | |||||
| 1 | 10 | Right | 14.13 | R | Lower | X | |||||
| 1 | 11 | 13.81 | NR | Upper | X | ||||||
| 1 | 25 | 14.14 | R | Upper | X | ||||||
Note. DL = dichotic listening; Post = Posterior; Ant = Anterior; CC = corpus callosum; AC = anterior commissure; Hand = handedness; NR = non-righthanded
References
- Alexander AL, Lee JE, Lazar M, & Field AS (2008). Diffusion tensor imaging of the brain. Neurotherapeutics, 4(3), 316–329. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Anderson V, Northam E, Hendy J, & Wrennall J (2001). Developmental neuropsychology: A clinical approach. New York: Psychology Press. [Google Scholar]
- Arrington CN, Ware AL, Ahmed Y, Kulesz PA, Dennis M, & Fletcher JM(2016). Are shunt revisions associated with IQ in congenital hydrocephalus? A meta-analysis. Neuropsychology Review, 26, 329–339. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Beery K (1982). Revised administration, scoring, and teaching manual for the Development of Visual-Motor Integration. New York, NY: Modern Curriculum Press. [Google Scholar]
- Bryden MP (1988). An overview of the dichotic listening procedure and its relation to cerebral organization. In Hugdahl K(Ed.), Handbook of dichotic listening: Theory, methods, and research (pp. 1–43). Chichester: Wiley. [Google Scholar]
- Bradley KA, Juranek JJ, Romanowska-Pawliczek A, Hannay HJ, Cirino PT, Dennis M, Kramer LA, & Fletcher JM (2016). Plasticity of interhemispheric temporal lobe white matter pathways due to early disruption of corpus callosum development in spina bifida. Brain Connectivity, 6(3), 238–248. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Copp AJ, Adzick NS, Chitty LS, Fletcher JM, Holmbeck GN, & Shaw GM (2015). Spina bifida. Nature Review Disease Primers, 1, 1–18. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Dennis M, Landry SH, Barnes M, & Fletcher JM (2006). A model of neurocognitive function in spina bifida over the life span. Journal of the International Neuropsychological Society, 12, 285–296. [DOI] [PubMed] [Google Scholar]
- Dewan MC, & Wellons JC (2019). Fetal surgery for spina bifida: JNSPG 75th anniversary invited review article. Journal of Neurosurgery: Pediatrics, 24(2), 105–114. [DOI] [PubMed] [Google Scholar]
- Fischer M, Ryan SB, & Dobyns WB (1992). Mechanisms of interhemispheric transfer and patterns of cognitive function in acallosal patients of normal intelligence. Archives of Neurology, 49, 271–277. [DOI] [PubMed] [Google Scholar]
- Fletcher JM, Copeland K, Frederick J, Blaser SE, Kramer LA, Northrup H, Hannay HJ, Brandt ME, Francis DJ, Villarreal G, Drake JM, Laurent J, Townsend I, Inwood S, Boudousquie A & Dennis M (2005). Spinal lesion level in spina bifida: A source of neural and cognitive heterogeneity. Journal of Neurosurgery: Pediatrics, 102, 268–279. [DOI] [PubMed] [Google Scholar]
- Hannay HJ (2000). Functioning of the corpus callosum in children with early hydrocephalus. Journal of the International Neuropsychological Society, 6, 351–361. [DOI] [PubMed] [Google Scholar]
- Hannay HJ, Dennis M, Kramer L, Blaser S, & Fletcher JM (2009). Partial agenesis of the corpus callosum in spina bifida meningomyelocele and potential compensatory mechanisms. Journal of Clinical and Experimental Neuropsychology, 31(2), 180–194. doi: 10.1080/13803390802209954 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hannay HJ, Walker A, Dennis M, Kramer L, Blaser S, & Fletcher JM (2008). Auditory interhemispheric transfer in relation to patterns of partial agenesis and hypoplasia of the corpus callosum in spina bifida meningomyelocele. Journal of the International Neuropsychological Society, 14, 771–781. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hollingshead A (1975). Four factor index of social status. New Haven, CT: Yale University, Department of Sociology. [Google Scholar]
- Huber-Okrainec J, Blaser SE, & Dennis M (2005). Idiom comprehension deficits in relation to corpus callosum agenesis and hypoplasia in children with spina bifida meningomyelocele. Brain and Language, 93(3), 349–368. doi: 10.1016/j.bandl.2004.11.002 [DOI] [PubMed] [Google Scholar]
- Javad F, Warren JD, Micallef C, Thornton JS, Golay X, Yousry T, & Mancini L (2014). Auditory tracts identified with combined fMRI and diffusion tractography. NeuroImage, 84, 562–574. doi: 10.1016/j.neuroimage.2013.09.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Jenkinson M, Beckman CF, Behrens TE, Woolrich MW, & Smith SM (2012).FSL. NeuroImage, 62, 782–790. [DOI] [PubMed] [Google Scholar]
- Jones DK (2011). Diffusion MRI: Theory, Methods and Applications. New York, NY:Oxford. [Google Scholar]
- Kimura D (1961). Cerebral dominance and the perception of verbal stimuli. Canadian Journal of Psychology, 15, 166–171. [Google Scholar]
- Rosenzweig MR (1951). Representations of the two ears at the auditory cortex.American Journal of Physiology, 167, 147–158. [DOI] [PubMed] [Google Scholar]
- Satz P (1977). Laterality tests: An inferential problem. Cortex, 13, 208–212. [DOI] [PubMed] [Google Scholar]
- Siffredi V, Anderson V, Leventer RJ, & Spencer-Smith MM (2013).Neuropsychological profile of agenesis of the corpus callosum: a systematic review. Developmental Neuropsychology, 38(1), 36–57. doi: 10.1080/87565641.2012.721421 [DOI] [PubMed] [Google Scholar]
- Smith SM, Jenkinson M, Woolrich MW, Beckman CF, Behrens TE,Johansen-Berg H, Bannister PR, De Luca M, Drobnjak I, Flitney DE, &Matthews PM (2004). Advances in functional and structural MR image analysis and implementation as FSL. NeuroImage, 23(S1), 208–219. [DOI] [PubMed] [Google Scholar]
- Springer SP (1986). Dichotic listening. In Hannay HJ(Ed.), Experimentaltechniques in human neuropsychology. New York: Oxford University Press. [Google Scholar]
- Thorndike RL, Hagen EP, & Sattler JM (1986). Stanford-Binet Intelligence Scale:Fourth Edition. Chicago: Riverside. [Google Scholar]
- Voyer D (2011). Sex differences in dichotic listening. Brain and Cognition, 76(2), 245255. [DOI] [PubMed] [Google Scholar]
- Wahl M, Strominger Z, Jeremy RJ, Barkovich AJ, Wakahiro M, Sherr EH, &>Mukherjee P (2009). Variability of homotopic and heterotopic callosalconnectivity in partial agenesis of the corpus callosum: A 3T diffusion tensor imaging and Q-ball tractography study. American Journal of Neuroradiology, 30, 282–289. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Westerhausen R, Gruner R, Specht K, & Hugdahl K (2009). Functional relevanceof interindividual differences in temporal lobe callosal pathways: A DTI tractography study. Cerebral Cortex, 19, 1322–1329. [DOI] [PubMed] [Google Scholar]
- Wheeler-Kingshott C, & Cercignani M (2009). About “axial” and “radial” diffusivity.Magnetic Resonance in Medicine,61, 1255–1260. [DOI] [PubMed] [Google Scholar]
- Woolrich MW, Jbabdi S, Patenaude B, Chappell M, Makni S, Behrens T,Beckmann C, Jenkinson M, & Smith SM (2009). Bayesian analysis of neuroimaging data in FSL. Neuroimage, 45(1 Suppl):S173–86. [DOI] [PubMed] [Google Scholar]