Magnetic Resonance Imaging

Objective. Quantitative T2 mapping plays a critical role in brain imaging for assessing a range of neurological conditions, including neurodegenerative diseases, demyelinating disorders, and cerebrovascular pathologies. Despite its diagnostic potential, implementing quantitative T2 mapping at ultra-high magnetic field strengths ([≥]7T) poses significant challenges. These include elevated specific absorption rate (SAR) and radiofrequency (RF) field inhomogeneities, which can lead to prolonged scan durations and inaccuracies in quantification. Materials and Methods. Phase-based gradient-recalled echo (GRE) techniques have recently emerged as promising rapid acquisition with enhanced sensitivity to T2-related contrast. In this study, we introduce TWISTARE (TWo Interleaved Steady-states for T2 and RF Estimation), a novel dual steady-state 3D-GRE approach that employs interleaved flip angles and small RF phase increments to jointly estimate T2 and B1 maps. By combining two dual-steady-state scans, TWISTARE enables fast, whole-brain quantitative T2 mapping while reducing scan time and mitigating B1-related bias at ultra-high field. Results. Validation experiments included Bloch simulations, phantom studies and in-vivo imaging. The results demonstrated high precision in phantom experiments, achieving up to a two-fold reduction in acquisition time and achieved precision comparable to the gold-standard method in vivo within a similar scan duration. Discussion. TWISTARE establishes a fast steady-state framework for quantitative neuroimaging at ultrahigh field, offering potential benefits for both clinical and research applications, especially in longitudinal and dynamic studies of brain tissue.

Brain age is a promising biomarker for detecting atypical and pathological brain aging, but its accuracy and reliability depend critically on MRI quality. The impact of common MR image degradations such as motion, ghosting, blurring, and noise on brain age predictions remains unclear. In this study, we systematically assessed the effects of four simulated MRI artifact types, across ten severity levels, on brain age prediction using three widely used deep learning-based algorithms (Pyment, MIDI, MCCQR), in high-quality T1-weighted images of healthy adults (age range 18-85, 54% female). Artifact severity levels (1-10) were generated using a power-function mapping of TorchIO simulation parameters calibrated to the full PondrAI QC visual rating scale (from perfect to severely degraded image quality). Linear mixed-effects models with predicted brain age as dependent variable revealed a significant interaction between algorithm, artifact type, and severity (p<0.001), indicating algorithm-specific sensitivity to artifacts. In artifact-free scans, mean absolute error (MAE) was 4.6 years for MCCQR, 7.1 years for Pyment, and 9.1 years for MIDI. At severity level 10, MAE increased with up to 110% for Pyment, 112% for MCCQR, and 16% for MIDI (motion); and with up to 75% for Pyment, 135% for MCCQR, and 34% for MIDI (ghosting). Blurring had minimal impact at low-moderate levels, but at maximum severity MAE increased by 26% for Pyment and 137% for MCCQR, while MIDI remained largely stable. Noise minimally affected Pyment and MCCQR (MAE increases [&le;]20%), but led to larger declines for MIDI (MAE increase 35%). The vulnerability of different algorithms highlights that training data, preprocessing strategies and underlying architectures influence robustness, emphasizing that artifact sensitivity is a key consideration when interpreting brain-age as a biomarker. Our results emphasize the need for artifact-aware evaluation and mitigation strategies when algorithms such as brain age are used in clinical research.

Despite the increased water content in fibrotic livers, numerous studies reported a decrease in ADC (apparent diffusion coefficient) in liver fibrosis. We argue that the ADC decrease in fibrotic livers is due to the T2 shine-through of ADC, as the longer T2 in liver fibrosis leads to less signal decay between the low and high b-value images. The metric slow diffusion coefficient (SDC) was proposed to mitigate the difficulties associated with this T2 shine-through of ADC. This study calculated ADC and SDC of one rat study with liver fibrosis induced by biliary duct ligation (BDL), and three sets of human liver fibrosis data. To tease out the menopausal effect on SDC, only the results of mens livers were analysed for the human datasets. The rat study showed, liver ADC decreased stepwise (in weeks after BDL procedure) following fibrosis induction, SDC increased stepwise. In human studies, all three datasets consistently showed advanced fibrosis had an ADC lower than that of earlier stage fibrosis; advanced fibrosis had a SDC higher than that of earlier stage fibrosis. When each liver SDC datum was normalized by the mean value of the controls without fibrosis, and the three human datasets were summed together, stage-1 liver fibrosis had a normalized SDC value lower than that of the controls, and there was a stepwise increase of SDC value from stage-1 liver fibrosis to stage-4 liver fibrosis. It is known that liver fibrosis is associated with lower perfusion, higher iron/susceptibility, and higher water content, and these three factors all contribute to the lower ADC measure. Higher iron/susceptibility lowers SDC measure, whereas higher water content elevates SDC measure. It is likely that for early-stage fibrosis, the net effect of susceptibility and water leads to a lower SDC, while for advanced fibrosis, the net effect leads to a higher SDC.

IntroductionPreterm birth is a major risk factor for disrupted brain development and subsequent neurodevelopmental disorders, yet the underlying mechanisms remain poorly understood. Further, typical neuroimaging analyses are particularly challenging in the neonatal brain: data is frequently low quality and a lack of cellular development violates the assumptions relied on by many commonly-used techniques. In this study, we develop and present an advanced diffusion magnetic resonance imaging method to examine the microstructural organization of white matter in a clinically-acquired cohort of premature neonates. MethodsUsing a novel approach that resolves multiple tissue compartments within the brain, we provide highly detailed orientation and quantification of white matter fibers and tissue signal fraction. We also utilize a series of automated segmentation algorithms to identify and measure these metrics across key tracts and subcortical regions. We investigate how these measures relate to postmenstrual age, as well as to clinical factors reflecting neonatal illness severity. ResultsWe report successful segmentation and reconstruction of numerous white matter tracts throughout the neonatal brain. We further demonstrate the utility and functionality of microstructural analysis in a variety of pathologies commonly encountered in the neonatal clinical environment. Our results demonstrate tract-specific developmental trajectories, with early-maturing pathways showing higher microstructural organization. Exploratory analyses suggest that neonatal illness severity has modest, tissue-specific associations with microstructural properties. DiscussionThis work demonstrates that advanced microstructural imaging methods can extract meaningful white matter measurements from clinically-acquired scans, providing a practical framework for studying neonatal brain development in real-world hospital settings. These metrics are able to be calculated at extremely young ages, potentially allowing non-invasive study of vulnerable populations before detailed behavioral or neurological assessments are feasible.

Consistent, high-quality data is key to the success of fMRI studies given the many confounding factors and undesired signals that contaminate these data. Several quality assurance (QA) metrics exist for fMRI (e.g., temporal signal-to-noise ratio (TSNR), percent ghosting, motion estimates), but none of them leverage relationships between echoes that are part of multi-echo (ME) fMRI acquisitions. Here, we fill this gap by proposing a new QA metric for for ME-fMRI that quantifies the likelihood a given ME scan is dominated by BOLD (Blood Oxygenation Level-Dependent) fluctuations. We refer to this metric as pBOLD; the probability of the signal change being primarily BOLD contrast-dominated. Having an estimate of overall BOLD weighting - both before and after preprocessing - is meaningful because BOLD is the intrinsic contrast mechanism used in fMRI to infer neural activity. We introduce pBOLD to the neuroimaging community by first describing the theoretical principles supporting the metric. Next, we validate pBOLD efficacy using a small dataset (N=7 scans) of constant- and cardiac-gated scans that have distinct levels of contributing BOLD fluctuations. Third, we apply pBOLD to a larger publicly available ME dataset (N=439 scans), to evaluate six different pre-processing pipelines, and show how pBOLD provides complementary information to TSNR. Our results show that ME-based denoising increases both pBOLD and TSNR relative to basic denoising; however, including the global signal (GS) as a regressor only improves TSNR, but worsens pBOLD. Further analyses looking at the BOLD-like characteristics of the GS and its relationship to cardiac and respiratory traces suggest that the observed decrease in pBOLD is likely due to a decrease in BOLD fluctuations of neural origin contributing to the GS, and not due to contributions from other physiological BOLD fluctuations (i.e., respiratory and cardiac function). Finally, we also demonstrate how pBOLD can be applied as a data quality metric, by showing how higher pBOLD results in better ability to predict phenotypes based on whole-brain functional connectivity matrices.

Different MRI image contrasts are designed to highlight various tissue properties and combining them allows extension of probabilistic segmentation beyond the commonly used "gray-white-CSF" models. This work describes a fully automated method that combines T1-weighted, T2-FLAIR, and conventional T2-weighted images to provide internal consistency across prediction of tissue segmentations including segmentation of superficial and deep gray matter, white matter hyperintensities, and MR-visible perivascular spaces. Results from 773 imaging datasets from 403 participants in the Mayo Clinic Study of Aging and Mayo Clinic Alzheimers Disease Research Center (ADRC) are presented.

PURPOSEGlycolytic production of HDO from the metabolism of perdeuterated glucose provides a means for metabolic imaging with 2H MRI. The present study compared HDO production from a cost-efficient [2,3,4,6,6-2H5]glucose with [2H7]glucose in vitro and in vivo. METHODS2H NMR spectroscopy was performed to measure glucose consumption, lactate, and HDO production in the SFxL glioblastoma cell line. In vivo studies in healthy mice using 2H magnetic resonance spectroscopy were performed at 11.1 T after administering a bolus of either metabolic contrast agent. In vivo metabolite levels were quantified using unlocalized and slice-selective localized spectra. RESULTSOur in vitro results demonstrated similar glucose consumption and HDO production kinetics, although significant differences in lactate labeling were observed. The in vivo study showed comparable glucose consumption and HDO production kinetics following tail-vein bolus administration of either metabolic contrast agent, while lactate was not detected in the brain. CONCLUSION[2,3,4,6,6-2H5]glucose shows comparable HDO production to [2H7]glucose, while offering lower cost and reduced spectral complexity. These findings place [2,3,4,6,6-2H5]glucose as an alternative to [2H7]glucose for HDO-based DMI studies.

PurposeTo create a realistic in-silico brain phantom for positive and negative magnetic susceptibility that incorporates susceptibility anisotropy, enabling the evaluation of how susceptibility anisotropy influences susceptibility separation algorithm performance. MethodsWe expanded an existing QSM validation phantom by creating separate maps for positive and negative susceptibility, with the option of modeling susceptibility anisotropy. Multi-echo gradient echo data were simulated to evaluate four susceptibility separation techniques ({chi}-separation, DECOMPOSE-QSM, APART-QSM, and [Formula]). To assess the impact of noise, simulations were performed at different SNR levels (50, 100, 200, 300). ResultsOur findings showed that the error in negative susceptibility estimates increased by up to 53% when susceptibility anisotropy was present, compared to the case without susceptibility anisotropy, with {chi}-separation being the algorithm that was most sensitive to anisotropy. Robustness to noise varied across the assessed algorithms, with APART-QSM and {chi}-separation having the highest and lowest sensitivity to noise, respectively. ConclusionThe modified phantom is open-source and can serve as a numerical ground truth for evaluating susceptibility separation methods. Our findings emphasize the importance of incorporating susceptibility anisotropy into susceptibility separation models to improve their accuracy.

BACKGROUNDMild traumatic brain injury (mTBI) is a signature injury in civilian and military populations that remains invisible to detection by conventional radiological methods. Diffusion MRI has been identified as a potential clinical tool for revealing subtle microstructural alterations associated with mTBI. OBJECTIVEThis study evaluates whether a comprehensive and powerful diffusion MRI (dMRI) technique called mean apparent propagator (MAP) MRI can detect sequelae of mTBI. METHODSWe analyzed data from 417 participants of the GE/NFL prospective mTBI study which included 143 matched controls (mean age, 21.9 {+/-} 8.3 years; 76 women) and 274 patients with acute mTBI and GCS [&ge;]13 (mean age, 21.9 {+/-} 8.5 years; 131 women). All participants underwent MRI exams at up to four visits including structural high-resolution T1W, T2W, FLAIR-T2W, and dMRI, in addition to clinical assessments of post-concussive physical symptoms (RPQ-3), psychosocial functioning and lifestyle symptoms (RPQ-13), and postural stability (BESS). The dMRI data for each subject were co-registered across all visits and analyzed using the MAP-MRI framework to measure and map the distribution of net microscopic displacements of diffusing water molecules in tissue and ultimately compute the microstructural MAP-MRI tissue parameters including propagator anisotropy (PA), Non-Gaussianity (NG), return-to-origin probability (RTOP), return-to-axis probability (RTAP), and return-to-plane probability (RTPP). We quantified voxel-wise and region-of-interest (ROI)-based changes in these parameters across all four visits. RESULTSMAP-MRI parameter values were within the expected ranges and showed relatively little variation across visits. We found no significant differences in the longitudinal trajectories of these parameters between mTBI patients and controls. At acute post-injury timepoints, RPQ-3 and RPQ-13 scores were increased in mTBI patients relative to controls, while BESS scores were not significantly different between groups. Analysis of dMRI metrics and clinical mTBI markers showed significant correspondence between MAP-MRI metrics in cortical gray matter, caudate and pallidum and BESS scores. CONCLUSIONWe developed and tested a state-of-the-art quantitative image processing pipeline for sensitive analysis and detection of subtle tissue changes in longitudinal clinical diffusion MRI data. The absence of a significant statistical difference between populations in the dMRI parameters in this study suggests that the mTBI corresponded to acute post-injury clinical symptoms but that the injury was not severe enough to cause detectable microstructural damage/alterations, and that increased diffusion sensitization combined with improved analysis techniques may be needed. CLINICAL IMPACTThese findings suggest that acute mTBI (GCS[&ge;]13) may not be detectable with diffusion MRI. TRIAL REGISTRATIONClinicalTrials.gov NCT02556177

Diffusion-weighted magnetic resonance imaging (dMRI) is used to study white matter microstructure and to delineate pathways by estimating fiber orientation distributions (FODs). Symmetric FODs represent the conventional model assuming antipodal symmetry in water diffusion. However, in complex regions with bending, branching or fanning fibers, this assumption is not guaranteed. To better capture such underlying fibers geometries, asymmetric FODs (A-FODs), derived from neighboring FODs, have been introduced. Here, we propose an Encoder-based Curvature-Aware Regularization (EnCAR) method for estimating A-FODs. Incorporating curvature features into the regularization weight applied to neighboring voxels improves reconstruction of A-FODs. A self-supervised Transformer network, combined with a Spherical Harmonics Semantic Encoder, learns region-specific regularization parameters from this local neighborhood to capture the diversity of fiber geometries across the brain. The EnCAR method was verified on the DiSCo challenge phantom, and applied to in vivo multi-shell Human data. The model estimated sharp, high-angular-resolution A-FODs that were well aligned with local fiber pathway. Compared with established FOD and A-FOD methods, it performed on par in regions dominated by symmetric FODs and outperformed them in complex asymmetric regions. Quantitative evaluation using the Asymmetry Index (ASI) and Model Discrepancy Index (MDI) confirmed improved consistency with the underlying diffusion signals. By ensuring smooth directional transitions, this work enhances the visibility of continuous fiber segments.

Compared with model-based and phenomenological descriptions of the spatiotemporal dynamics of resting-brain activity, statistical characterizations of resting-state fMRI (rs-fMRI) data remain relatively underexplored. Some sophisticated analysis techniques, such as Mapper-based topological data analysis (TDA) and innovation-driven coactivation pattern analysis (iCAP), can distinguish real data from phase-randomized (PR) surrogates, suggesting that rs-fMRI data are not as simple as stationary Gaussian processes. However, the exact statistical properties that distinguish real rs-fMRI data from PR surrogates have not yet been determined. In this study, we conducted system identification analysis and surrogate data analysis to specify key statistical properties that allow TDA and iCAP to discriminate real rs-fMRI data from PR surrogates. We first analyzed rs-fMRI data concatenated across scans using autoregressive (AR) modeling and found that the scan-concatenated rs-fMRI data were weakly non-Gaussian. However, non-Gaussianity alone was insufficient to reproduce realistic TDA and iCAP results because of non-stationarity across scans. AR modeling of single-scan data revealed that rs-fMRI data were statistically indistinguishable from a Gaussian distribution within a single scan, although TDA and iCAP results still differed between the real data and PR surrogates. A new surrogate dataset designed to preserve non-stationarity successfully reproduced realistic TDA and iCAP results, suggesting that TDA and iCAP likely capture the non-stationarity of rs-fMRI data to distinguish it from PR surrogates. Together, these results indicate approximate Gaussianity and non-stationarity in rs-fMRI data, providing a data-driven and statistical characterization of resting-state brain activity that can serve as a quantitative reference for whole brain simulations and generative models.

Purpose: To develop a deep neural network-based, AIF-free, perfusion estimation method (QTMnet) for improved performance on glioma classification. Methods: A globally defined arterial input function (AIF) is needed to recover perfusion parameters in the two-compartment exchange model (2CXM). We have developed Quantitative Transport Mapping (QTM) to create an AIF-independent estimation method. QTM estimation can be formulated using deep neural networks trained on synthetic DCE-MRI data (QTMnet). Here, we provide a fluid mechanics-based DCE-MRI simulation with exchange between the capillaries and extravascular extracellular space. We implemented tumor ROI generation to morphologically characterize tissue perfusion. We compared our QTMnet implementation with 2CXM on 30 glioma human subjects, 15 of which had low-grade gliomas, and 15 with high-grade glioblastomas. Results: QTMnet outperforms (best AUC: 0.973) traditional 2CXM (best AUC: 0.911) in a glioma grading task. Conclusion: The AIF-independent QTMnet estimation provides a quantitative delineation between low-grade and high-grade gliomas.

Mechanical properties of biological tissues vary across spatial scales, yet radiomics typically relies on fixed, heuristic choices for neighbourhood size, kernel geometry, and spectral content - choices that can silently reshape the feature space before any modelling begins. We introduce a label-free, information-theoretic framework for selecting extraction parameters in multi-frequency MRE radiomics. For each configuration {theta} - neighbourhood radius r, kernel geometry k (sphere or shell), and frequency subset f - we extract a radiomics feature matrix and score it using an objective J({theta}) that integrates distributional richness (Shannon entropy), cross-frequency coherence (canonical correlation), inter-feature redundancy (Spearman correlation), and bootstrap stability. We evaluate 121 configurations per tissue in multi-frequency MRE (30-60 Hz) of human brain, liver, and a calibrated phantom, and test robustness using 10,000 Dirichlet-sampled objective weightings. Across tissues, neighbourhood aggregation is consistently preferred over voxel-wise extraction, outperforming the no-neighbourhood baseline in 98.4-100% of weightings. External validation in 100 independent brain scans acquired with a different protocol and wider frequency range (20-90 Hz) confirms a reproducible mesoscopic plateau at r = 3-5 (9-15 mm), with a modal optimum at r = 4; omitting neighbourhood analysis reduces J({theta}) by 38% relative to each subject's optimum. Frequency-subset preferences replicate across datasets, with lower frequencies most frequently selected for brain. By turning ad hoc extraction choices into an outcome-free optimisation step, this framework improves reproducibility, reduces sensitivity to heuristic parameter choices, and generalises across acquisition protocols and imaging sites.

Resting-state functional MRI has been widely used to study brain connectivity, yet the test-retest reliability of commonly used metrics remains a concern. To improve reliability, extended scan lengths and larger subject cohorts are often recommended. However, these solutions can be impractical and pose challenges, particularly in studies of clinical populations. Here, we systematically assess the reliability of two main types of functional connectivity measures: node-based connectome metrics (edge-level intraclass correlation coefficient [ICC], connectome-level ICC, functional connectivity fingerprinting, and discriminability); and voxel-based resting-state networks (RSNs) (spatial similarity of independent component analysis [ICA]-derived RSN maps quantified using the Dice coefficient). Using data from the Human Connectome Project, we evaluated the effects of scan length (3.6, 7.2, 10.8, and 14.4 minutes) and number of participants (n = 10, 20, 50, and 100), on both within-session and between-session reliability. We found that multivariate connectome metrics demonstrated greater reliability than edge-level measures, and that scan length had stronger influence on test-retest reliability than the number of participants. For connectome metrics, 14 minutes of scanning and a cohort of approximately 20 participants were sufficient to achieve reliable estimates. In contrast, RSN measures benefited from larger cohort sizes. Our findings provide practical guidelines for designing resting-state fMRI studies in terms of scan length and number of participants, balancing reliability and feasibility. Ultimately, protocol choices should be guided by the specific study objectives and the functional connectivity metric of interest.

Functional brain mapping is an important tool to understand the organization of the human brain, both at the group level, but also to an increasing degree at the level of the individual. There are currently two main approaches to do so. Resting-state fMRI relies on inter-regional correlations of random fluctuations of the signal. In contrast, task-based localizers typically use a single-contrast between a task of interest and a matched control task to identify the location of a functional region in an individual brain. In this paper, we propose and evaluate a third approach: the use of multi-task batteries for both localization of a single functional region and parcellation of multiple functional regions. We show that multi-task localizers produce more consistent estimation of a single functional region across subjects than the single-contrast approach using the same amount of fMRI data. Furthermore, we demonstrate that the multi-task approach is sensitive to true inter-individual differences in region size, and does not suffer the same influence of signal-to-noise ratio that biases the single-contrast localizer. We then address the question of how to select tasks for the battery, and present a data-driven strategy that optimizes the characterization of a brain structure of interest. We show that such batteries outperform randomly selected batteries both for building individual parcellations as well as individual connectivity models. Finally, we demonstrate that an interspersed design - where all tasks are presented in each imaging run - yields more reliable results than splitting the tasks across different runs. We present an open source toolbox for the implementation of multi-task batteries, along with a library containing group-averaged activity patterns that can be used to optimize battery selection for different brain structures of interest.

Cerebrovascular reactivity (CVR), the ability of cerebral blood vessels to dilate or constrict in response to a vasoactive stimulus, is an important measure of cerebrovascular health. Accurate CVR estimation requires accounting for the time required for the vasoactive stimulus to reach each brain region and the time it takes for local arterioles to modulate cerebral blood flow. The temporal search range used to calculate this spatially varying offset can substantially impact CVR estimates, and the appropriate search range may vary across populations, acquisition protocols, and even brain regions. Here, we present an iterative approach for automatically determining the appropriate maximum shift, using breath-hold fMRI data acquired in a cohort of stroke survivors. This approach selectively expands the delay search range only for voxels with estimated delays at the boundary (i.e., near the minimum or maximum shift) until the estimated delay is no longer constrained or a predefined value is reached. In the context of stroke, this approach significantly increased the number of voxels with statistically significant CVR among those initially at the boundary. It also resulted in CVR polarity reversals in voxels originally at the early-response boundary and amplified negative CVR values in voxels originally at the late-response boundary, suggesting that using an iterative maximum shift can critically impact CVR interpretation. This approach is broadly applicable beyond stroke, but careful parameter tuning is required, as illustrated by our demonstration of the parameter tuning process for a participant with Moyamoya disease. Together, these findings suggest that iterative delay correction allows for improved CVR assessments in clinical populations.

The hippocampal CA1 subregion supports learning, memory formation, and spatial navigation. Although its three-layered architecture has been described in ex-vivo investigations, the in-vivo microstructural profile of CA1 and its relation to individual variations in memory performance remain poorly characterized. In this study, we used ultra-high field structural MRI at 7 Tesla to investigate the depth-dependent myelination patterns (measured by quantitative T1) of CA1 in younger adults, their relation to the local arterial architecture, and their association with individual differences in cognitive functions, specifically memory performance. Results show that left and right CA1 present depth-dependent patterns of myelination, with the outer and inner compartments showing higher myelination than the middle compartment. No significant relationship between layer-specific myelination of CA1 and distance to the nearest artery was observed. Right CA1 was found to be more myelinated than left CA1. Pairwise correlations and regression models showed that higher left CA1 myelination is linked to higher accuracy in object localization. Together, our data demonstrates the feasibility of describing the three layered myelin architecture of CA1 in vivo, and provides information on how alterations in the architecture of CA1 may relate to alterations in cognitive performance in younger adults.

Foundation models (FMs) for brain magnetic resonance imaging (MRI) are increasingly adopted as pretrained backbones for clinical tasks such as brain age prediction, disease classification, and anomaly detection. However, if FM embeddings (internal representations) shift systematically across MRI scanners, downstream analyses built on them may reflect acquisition hardware rather than biology. No study has yet quantified this cross-scanner reproducibility. Here, we assess the cross-scanner reliability of brain MRI FM embeddings and investigate which design factors (pretraining strategy, network architecture, embedding dimensionality, and pretraining dataset scale) best explain the observed differences. Using the ON-Harmony travelling-heads dataset (20 participants, eight scanners, three vendors), we evaluate the embeddings of five architecturally diverse FMs and a FreeSurfer morphometric baseline via within- and between-scanner intraclass correlation coefficient (ICC), variance decomposition, and scanner fingerprinting. Reliability spanned the full spectrum: biology-guided models achieved good-to-excellent cross-scanner ICC (AnatCL: 0.970 [95\% confidence interval (CI): 0.94, 0.98]; y-Aware: 0.809 [0.63, 0.88]), matching or surpassing FreeSurfer (0.926 [0.83, 0.96]), whereas purely self-supervised models fell below the poor threshold (BrainIAC: 0.453, BrainSegFounder: 0.307, 3D-Neuro-SimCLR: 0.247), with 23--58\% of embedding variance attributable to scanner identity. The strongest correlate of cross-scanner reliability among the models evaluated was pretraining strategy: incorporating biological metadata (cortical morphometrics, age) into the contrastive objective produced scanner-robust embeddings, whereas architecture, dimensionality, and dataset scale did not predict reliability.

Global Signal Removal (GSR) is a widely applied step in functional magnetic resonance imaging (fMRI) preprocessing. Although GSR conventionally denotes Global Signal Regression, we use Global Signal Removal to encompass a broader family of spatial filtering operations. GSR in general remains controversial due to concerns about introducing spurious anticorrelations and removing neurally meaningful signals. In this paper, we provide a precise geometric characterization by formalizing GSR as graph spatial filtering. We demonstrate that the most common form of GSR, Regression-GSR, equates to a rank-1 deflation of the covariance matrix (i.e. functional connectivity) by the degree vector. Empirically, the degree vector is dominated by the first principal component of the functional connectivity matrix (correlation = 0.88 {+/-} 0.12 in resting-state HCP data), making Regression-GSR an approximation to first eigenmode removal. This view of GSR as a spatial projection framework allows us to develop a family of GSR variants, each expressible in a unified spatial filter: Naive-GSR removes the uniform vector, PCA-GSR precisely removes the first eigenvector, and SC-GSR, a new variant we introduce that removes the first harmonic of the structural connectivity matrix. A key distinction emerges: while Naive, PCA, and SC-GSR are orthogonal projections, Regression-GSR is an oblique projection that computes regional weights proportional to the degree vector but removes a spatially uniform signal. All GSR variants induce numerical singularity in the covariance matrix, but they differ in their effects on task-state separability, which we examine empirically. In summary, we reframe GSR as a family of graph spatial filters that enable interpretability of its effects, with systematically varying effects on network connectivity across variants.

Computer vision and deep learning techniques, including convolutional neural networks (CNNs) and transformers, have increased the performance of medical image classification systems. However, training deep learning models using medical images is a challenging task that necessitates a substantial amount of annotated data. In this paper, we implement data augmentation strategies to tackle dataset imbalance in the VinDr-SpineXR dataset, which has a lower number of spine abnormality X-ray images compared to normal spine X-ray images. Geometric transformations and synthetic image generation using Generative Adversarial Networks are explored and applied to the abnormal classes of the dataset, and classifier performance is validated using VGG-16 and InceptionNet to identify the most effective augmentation technique. Additionally, we introduce a hybrid augmentation technique that addresses class imbalance, reduces computational overhead relative to a GAN-only approach, and achieves ~99% validation accuracy with both classifiers across all three case studies. Keywords: Data augmentation, Generative Adversarial Network, VGG-16, InceptionNet, Class imbalance, Computer vision, Spine X-ray, Radiology.