Contemporary accounts of semantic cognition propose that conceptual knowledge is supported by a heteromodal conceptual store and controlled retrieval processes. However, it remains unclear how the neural basis of semantic control varies across modalities. Recent models of cortical organisation suggest that control networks are distributed along a unimodal-to-heteromodal cortical gradient, with the semantic control network (SCN) located in more heteromodal cortex than the domain-general multiple demand network (MDN). We used fMRI to examine how these networks respond to semantic control demands in visual and auditory tasks. Participants judged the semantic relatedness of spoken and written word pairs. On half of the trials, a task cue specified the semantic feature to guide retrieval; on the remaining trials, no such cue was given. The SCN showed greater activation when task knowledge was available, consistent with a role in the top-down control of semantic retrieval across modalities. In contrast, the MDN showed greater activation for spoken words, likely reflecting increased demands in speech perception. These findings demonstrate a dissociation between control networks, with SCN involvement modulated by task structure and MDN activity influenced by perceptual difficulty.

The relationship between the structural connectome and functional activity in the brain is highly complex, and understanding of the connection between the two is limited. Previous work has shown a marginal reliance of functional brain activity on underlying structural connections, indicating significant flexibility of neural communication. Here, we introduce a new method to quantify structure-function coupling and compare it with a standard coupling technique by evaluating the structure-function relationship across numerous fMRI task paradigms. Through this comparison, we investigate how structure-function relationships change during different cognitive demands and we evaluate how they relate to behavior. The new method introduced here, structural reliance, exhibits different structure-function correspondence patterns throughout the brain, and it generally outperforms the standard coupling measure in coupling-based behavioral measure predictions.

ABSTRACTPregnancy induces neuroanatomical changes in the human brain. Previous studies detected both traces of motherhood decades after childbirth and adaptations in fathers. It is unclear which effects can be attributed to persisting traces of pregnancy and which are effects of parenthood. We investigated effects of past birth and of pregnancy loss in women, and effects of fatherhood in men, using univariate and machine learning analyses on 205 regional brain volumes. A group of mothers and an age‐matched sample of nulliparous women (N = 4357 per group, mean age 63 years) from the UK Biobank, with no past pregnancy losses, showed significant volumetric group differences in 14 regions at Bonferroni‐adjusted α = 0.05. Likewise, we identified 18 significant group differences between age‐matched samples of fathers and non‐fathers of the same size (mean age 63.4), with 9 regions overlapping between sexes. Brain‐wide association statistics for past live birth in mothers and those for fatherhood correlated (r = 0.55). XGBoost machine learning models trained to classify parenthood status separately in both datasets showed performance that was low, but significantly above chance (10‐fold cross validation: AUC = 0.56, p < 1e‐5 Motherhood classifier, AUC = 0.54, p < 1e‐5, Fatherhood classifier, 10 k permutations). We tested the motherhood classification model on an independent test sample comprising four age‐matched groups: 1. women who have never been pregnant, 2. women with past pregnancy loss but no live births, 3. women with live births but no pregnancy loss, and 4. women who experienced both. Class probability was significantly associated with live births, but not past loss. These findings may suggest that neuroanatomical patterns of past childbirth partly also reflect traces of parenthood and not solely persisting traces of past pregnancy, although a more detailed characterization of pregnancy loss data would be needed for full confirmation of this interpretation. Therefore, further research is needed to quantify the extent and understand the nature of these changes, particularly considering the known vulnerability for mental disorders associated with reproductive events.

The reliance of fMRI research on between-person comparisons is limited by low test-retest reliability and an inability to explain within-person processes. Intraindividual studies are needed to understand how changes in brain functioning relate to changes in behavior. Here, we present open data and analysis of a novel intensively sampled fMRI study. This precision imaging dataset includes 44 sessions acquired across four participants at a twice-weekly rate. In each session, participants completed multiple reward-related tasks, mood and alertness ratings, and a behavioral mood manipulation. We examined how the reward response reflects between-person or within-person variance. Trial-level models suggested dramatically more trials than typically collected are needed to maximize reliability within runs and individuals. Test-retest reliability of the reward response was very low and not explained by measurement error, suggesting low power for between-person comparisons without large amounts of data. At an intraindividual level, mood and alertness explained up to 37% of the intraindividual variance of the anticipatory reward response. Finally, we found that while reliability or brain-behavior associations were not improved by multi-echo denoising, a multivariate reward signature had stronger intraindividual behavioral associations than a univariate anatomical mask. Together, results suggest that the BOLD reward response is not a stable trait-like marker, but moderated by state-like factors. More broadly, BOLD activation to reward tasks-and likely other fMRI tasks-presents substantial opportunity for within-person study to complement the traditional focus on between-person study. We conclude with a discussion of considerations for intensive longitudinal neuroimaging designs.

Face and scene perception rely on distinct neural networks centered on the Fusiform Face Area (FFA) and Parahippocampal Place Area (PPA). However, how these regions interact with broader brain networks remains unclear. Using resting-state fMRI and MEG data, we mapped the spatial and frequency-specific functional connectivity of the FFA and PPA. We found that the FFA showed predominant fMRI connectivity with lateral occipitotemporal, inferior temporal, and temporoparietal regions, while the PPA connected more strongly with ventral medial visual, posterior cingulate, and entorhinal-perirhinal areas. MEG analyses further revealed this network segregation was reflected in beta and gamma bands. Importantly, connectome-based predictive modeling showed that the strength of these intrinsic fMRI connectivity patterns predicted individual reaction times on corresponding face and scene perception tasks. Our findings demonstrate that the FFA and PPA anchor distinct intrinsic networks with unique spatio-temporal profiles that provide a functional architecture supporting their specialized roles in face and scene perception.

18F-FDG brain PET atlas is widely used in clinical and neuroscience research. However, variation in PET scanner technology can lead to misinterpretation and quantification errors with mismatched atlases. This study aimed to develop high-resolution, age-stratified FDG brain atlases harmonized across different effective image resolutions (EIRs) to support diverse PET/CT scanners. A total of 140 adults (aged 20-90 years, 72 males) were retrospectively enrolled and divided into seven age groups (n = 20). All underwent 18F-FDG PET/CT at China-Japan Friendship Hospital (Jan 2022-May 2023) using the Siemens Biograph Vision 600 scanner. PET scans (4 min) were acquired 60 min after 18F-FDG injection (4.44 MBq/kg). System EIR was measured using a Hoffman phantom, and images were smoothed to emulate 5.0-6.0 mm and 8.0-10.0 mm resolutions. Images were normalized to MNI space (1.0 mm) via ANTs, and subjects with outlier FDG uptake values were excluded. SUVRs were calculated using the whole cerebellum as reference. Age-stratified and EIR-harmonized atlases were created. Gray/white contrast (GWC), age correlations, and SUVR differences were analyzed. After quality control, 97 participants were included for atlas construction. Atlases for both EIRs were generated. GWC was higher at 5.0-6.0 mm than at 8.0-10.0 mm (e.g., 2.73 vs. 1.76 in 20-30 years). SUVRs in cortical and subcortical regions negatively correlated with age; brainstem SUVRs correlated only at lower resolution. SUVRs were higher at high resolution across regions. A set of age-dependent, harmonized FDG-PET atlases was developed and the quantitative metrics across age were characterized.

Amyotrophic lateral sclerosis (ALS) is a clinically heterogeneous neurodegenerative disease, and neuroimaging markers offer a promising avenue to capture this variability. Cortical folding alterations in ALS remain largely unexplored despite growing interest in neuroimaging markers of the disease. This study is the first whole-brain investigation of sulcal morphometry in ALS. A total of 222 individuals diagnosed with ALS and 194 healthy controls (HC) were recruited through the Canadian ALS Neuroimaging Consortium (CALSNIC) for a longitudinal investigation. Using T1-weighted MRI processed with FreeSurfer and BrainVISA, we extracted cortical thickness and four sulcal features: width, mean depth, surface area, and length, across 123 sulci. We identified widespread alterations in ALS, with 19 sulci showing increased width and length alongside reduced depth and surface area, primarily in frontal and parietal regions surrounding the motor strip. The central sulcus (CS) emerged as the most consistently affected region, displaying bilateral widening and reduced depth, changes that closely tracked motor decline. Longitudinal analyses revealed progressive widening and reduced depth and surface area of the CS. These alterations closely tracked the progression of motor symptoms over the course of the disease and aligned with regional cortical thickness alterations. Our findings demonstrate that brain folding patterns, and in particular CS, are altered in ALS and correlate with clinical progression, resembling the neurodegenerative pattern of the disease. By revealing complementary and sensitive changes, sulcal-based metrics may offer promising neuroimaging biomarkers for early detection, prognosis, and patient stratification in ALS.

Diffusion tensor imaging (DTI) corrected for the free-water (FW) enables the separation of a hindered Gaussian-like profile from an isotropic component, which represents diffusion found in cerebrospinal and interstitial fluids within the extracellular space of grey and white matter. The assessment of the reproducibility and reliability properties of FW-corrected DTI is a crucial factor in demonstrating the potential clinical utility of this refinement, particularly considering the examinations across multiple medical centres. This paper explores the variability, reliability, and separability properties of free-water volume fraction (FWVF) and FW-corrected DTI-based measures in healthy human brain white matter using publicly available test-retest databases acquired in (1) intra-scanner, (2) intra-scanner longitudinal and (3) inter-scanner settings under varying acquisition schemes. Three different estimation techniques to retrieve the FW-corrected DTI parameters tailored to single- or multiple-shell diffusion-sensitising magnetic resonance (MR) acquisitions are investigated: (i) a direct optimization of bi-tensor signal representation in the variational framework, (ii) the region contraction-based approach and (iii) the spherical means technique combined with a correction of diffusion-weighted MR signal prior to DTI estimation. We found the previous suggestion that the FW correction to DTI in a single-shell diffusion-weighted MR acquisition improves the repeatability of DTI-based measures may be data- and methodology-dependent, and does not generalise to multiple-shell scenarios. The study also confirms that the single-shell variational FW-correction method fails to retrieve meaningful information from the mean diffusivity (MD) parameter. In contrast, the combined FW-correction scheme reduces the biological variability of MD, regardless of whether DTI is estimated from single- or multiple-shell data, given that the FWVF used for the correction in both cases is derived from multiple-shell acquisitions. Our experiments have shown that the most reliable and repeatable/reproducible measures, while preserving a moderate separability property, are fractional anisotropy and axial diffusivity estimated in a multiple-shell variant under a combined FW-correction scheme. On the contrary, our results show evidence that the least reliable measures are the mean diffusivity estimated using any FW-correction procedure, as well as the FWVF parameter itself. These results can be used to establish the direction for selecting the most attractive FW-correction DTI scheme for clinical applications in terms of the variability-reliability-separability criterion.

Multimodal imaging studies that combine temporally and spatially precise methods can enhance understanding of reward neurocircuitry by revealing how signals specifically time-locked to reward processing substages relate to functional network-level activity. Prior reward studies that require decision making and/or motor responses to obtain rewards may complicate efforts to isolate basic reward responses from higher-order functions that support reward attainment. Here, we take an integrated, multimodal approach to evaluate anticipatory and consummatory reward processing substages, while minimizing demands on higher-order cognitive and motor processes. Functional magnetic resonance imaging (fMRI) and electrophysiological (EEG) data were separately recorded from 52 adults playing a simple, slot machine task, with reward outcomes independent of performance. Joint Independent Component Analyses (jICAs) were conducted with EEG-based event-related potential (ERP) difference waveforms and fMRI contrast images specific to reward anticipation and outcome processing substages. Resulting joint independent components (JICs) segregated reward processing substages, indicating significant co-modulation between temporal ERP and spatial fMRI signals (p < 0.001). During Reward Anticipation, a JIC with the temporal signature of the stimulus preceding negativity (SPN) ERP component covaried with fMRI activation in bilateral supplementary motor areas (pre-SMA/SMA) and inferior fronto-insular salience network regions implicated in attentional orienting and shifting. During Reward Feedback, JICs with the temporal signature of the reward positivity (RewP) ERP component covaried with fMRI activation in dorsal anterior cingulate cortex (dACC), ventral striatum, SMA, and inferior frontal cortex, extending to the insula. Further, trait reward sensitivity correlated with jICA-informed Win > Loss brain activations during Reward Feedback (p = 0.016). Our findings demonstrate that temporally precise electrophysiological and spatially rich hemodynamic measures of reward processing converge to map onto specific substages of reward-related brain processes. ERP and fMRI signaling during reward feedback support covariation of the RewP with dACC-striatal reward networks while the SPN covaried with fMRI signal in pre-SMA/SMA and inferior fronto-insular regions implicated in motor planning, salience, and attention.

Enlarged perivascular spaces (PVS) are imaging markers of cerebral small vessel disease (CSVD) that are associated with age, disease phenotypes, and overall health. Quantification of PVS is challenging but necessary to expand an understanding of their role in cerebrovascular pathology. Accurate and automated segmentation of PVS on T1-weighted images would be valuable given the widespread use of T1-weighted imaging protocols in multisite clinical and research datasets. We introduce segcsvdPVS, a convolutional neural network (CNN)-based tool for automated PVS segmentation on T1-weighted images. segcsvdPVS was developed using a novel hierarchical approach that builds on existing tools and incorporates robust training strategies to enhance the accuracy and consistency of PVS segmentation. Performance was evaluated using a comprehensive evaluation strategy that included comparison to existing benchmark methods, ablation-based validation, accuracy validation against manual ground truth annotations, correlation with age-related PVS burden as a biological benchmark, and extensive robustness testing. segcsvdPVS achieved strong object-level performance for basal ganglia PVS (DSC = 0.78), exhibiting both high sensitivity (SNS = 0.80) and precision (PRC = 0.78). Although voxel-level precision was lower (PRC = 0.57), manual correction improved this by only ~3%, indicating that the additional voxels reflected primary boundary- or extent-related differences rather than correctable false positive error. For non-basal ganglia PVS, segcsvdPVS outperformed benchmark methods, exhibiting higher voxel-level performance across several metrics (DSC = 0.60, SNS = 0.67, PRC = 0.57, NSD = 0.77), despite overall lower performance relative to basal ganglia PVS. Additionally, the association between age and segmentation-derived measures of PVS burden was consistently stronger and more reliable for segcsvdPVS compared to benchmark methods across three cohorts (test6, ADNI, CAHHM), providing further evidence of the accuracy and consistency of its segmentation output. segcsvdPVS demonstrates robust performance across diverse imaging conditions and improved sensitivity to biologically meaningful associations, supporting its utility as a T1-based PVS segmentation tool.