Most recent paper

Hum Brain Mapp. 2025 Aug 15;46(12):e70322. doi: 10.1002/hbm.70322.

ABSTRACT

Common measures of brain functional connectivity (FC) including covariance and correlation matrices are symmetry-positive definite (SPD) matrices residing on a cone-shaped Riemannian manifold. Despite its remarkable success for Euclidean-valued data generation, the use of standard generative adversarial networks (GANs) to generate manifold-valued FC data neglects its inherent SPD structure and hence the inter-relatedness of edges in real FC. We propose a novel graph-regularized manifold-aware conditional Wasserstein GAN (GR-SPD-GAN) for FC data generation on the SPD manifold that can preserve the global FC structure. Specifically, we optimize a generalized Wasserstein distance between the real and generated SPD data under adversarial training, conditioned on the class labels. The resulting generator can synthesize new SPD-valued FC matrices associated with different classes of brain networks, for example, brain disorder or healthy control. Furthermore, we introduce additional population graph-based regularization terms on both the SPD manifold and its tangent space to encourage the generator to respect the inter-subject similarity of FC patterns in the real data. This also helps in avoiding mode collapse and produces more stable GAN training. Evaluated on resting-state functional magnetic resonance imaging (fMRI) data of major depressive disorder (MDD), qualitative and quantitative results show that the proposed GR-SPD-GAN clearly outperforms several state-of-the-art GANs in generating more realistic fMRI-based FC samples. When applied to FC data augmentation for MDD identification, classification models trained on augmented data generated by our approach achieved the largest margin of improvement in classification accuracy among the competing GANs over baselines without data augmentation.

PMID:40820900 | PMC:PMC12358810 | DOI:10.1002/hbm.70322

Addict Biol. 2025 Aug;30(8):e70083. doi: 10.1111/adb.70083.

ABSTRACT

Adaptive stress coping is often impaired in individuals with alcohol use disorder (AUD). This process relies on neurocircuitry involved in emotional and behavioural regulation, particularly the ventromedial PFC (vmPFC) and orbitofrontal cortex (OFC), along with limbic and ventral striatal regions (e.g., amygdala, hippocampus and nucleus accumbens). These systems are highly sensitive to the neurotoxic effects of alcohol, which may disrupt their ability to flexibly adapt in response to acute stress. This study investigated state-dependent changes (termed 'flexibility') in vmPFC-limbic/striatal and OFC-limbic/striatal functional connectivity from rest to acute stress in individuals with AUD versus matched controls and examined associations with coping strategies. Twenty-four adults with AUD (agemean = 33, 11F) and 23 matched controls (agemean = 32, 11F) underwent fMRI during resting-state followed by the Montreal Imaging Stress Task (MIST) and completed the COPE Inventory. Functional connectivity between vmPFC-limbic/striatal and OFC-limbic/striatal regions was assessed during rest and stress (MIST) conditions. Group differences in state-dependent changes in functional connectivity were analysed using repeated-measures ANCOVA. Functional connectivity between the right OFC-right amygdala and right OFC-right hippocampus increased from resting-state to the MIST in the control group, but this shift was not present in the AUD group (group x condition, pFDR < 0.05). Although connectivity did not differ between groups during the MIST (p's > 0.2), the AUD group exhibited elevated connectivity between these regions at rest (p's < 0.05). Moreover, among controls, increased right OFC-right hippocampus connectivity from rest to MIST was associated with more adaptive versus maladaptive coping (p < 0.05). Compared to controls, individuals with AUD exhibited a pattern of inflexible OFC-amygdala and OFC-hippocampus functional connectivity under changing stress conditions. Diminished stress-related connectivity changes in AUD appeared to be driven by elevated functional connectivity at rest. Future studies should test whether this resting-state connectivity pattern reflects an allostatic state that constrains the system's capacity to flexibly respond to acute stress.

PMID:40820466 | PMC:PMC12358688 | DOI:10.1111/adb.70083