Magnetic resonance imaging, or MRI, is an important tool for finding and tracking spinal cord lesions in people with multiple sclerosis (MS), which are areas of damage caused by the disease. But automatic computer methods for detecting and outlining these lesions often work well only for one MRI type or one hospital’s scanning setup, which makes them less useful in real clinics where scan methods vary a lot. To address this, the researchers developed a more robust segmentation system, meaning a model that can automatically identify lesion boundaries, across many MRI contrasts and imaging sites. They trained and tested it on a large dataset of 4,428 annotated images from 1,849 people with MS across 23 imaging centers, using six different MRI contrast types and scans taken at 1.5, 3, and 7 tesla, which refers to the strength of the MRI scanner. Compared with existing methods that are designed for only one contrast type, the new model generalized better across different scan settings, according to neuroradiologist ratings. It also remained strong when tested across different spinal cord levels, image resolutions, threshold settings, and external datasets. Overall, the study shows that this approach can detect spinal cord MS lesions accurately and reliably across diverse MRI data, which is an important step toward making automated lesion analysis more useful in everyday clinical care.

Figure 1. Sankey diagram of annotated MRI scans across clinical sites. Line thickness is associated with the number of scans.

MRI scan distribution is clustered per acquisition type (3D, 2D sagittal, or 2D axial) and per MRI contrast, for each site.

Magnetic resonance imaging, or MRI, is often collected as a stack of 2D slices because that can make scans faster and improve image quality for clinical use. But when software tries to analyze these scans as if they were full 3D images, it can struggle, especially when the slices are thick or have gaps between them. To address this, the researchers developed ECLARE, a new method that improves the resolution of these slice-based MRI scans without needing outside training data. ECLARE first estimates the shape of each slice’s signal, then learns from the image itself how to turn lower-resolution parts into higher-resolution ones, while also correcting for blur and making sure the image is resampled in a way that respects the original field of view. The method was tested on brain MRI data, including healthy T1-weighted scans and T2-FLAIR scans from people with multiple sclerosis, and compared with several existing image-enhancement methods. Across scans with slice thicknesses up to 5 mm and gaps up to 1.5 mm, ECLARE produced more accurate and visually similar images than the alternatives, including in important brain regions such as the ventricles, caudate, and white matter. Overall, the study suggests that ECLARE can make thick-slice MRI images more useful for 3D analysis, which could improve downstream tools that rely on detailed brain structure.

Fig.��1

Flowchart of our proposed method. The anisotropic input volume is fed independently into each of the three steps. First, in panel a (Sec.��), we estimate the slice excitation profile with ESPRESO.��Second, in panel b (Sec.��), we extract HR in-plane 2D patches and use the PSF estimated from panel a to create paired training data. This training data are used to train the network��𝑓𝜃��with supervised learning. Third, in panel c (Sec.��), we extract LR through-plane 2D slices and superresolve them with the trained network��𝑓𝜃��from panel b. The superresolved slices are stacked and averaged, yielding the superresolved output volume.

Language comprehension is a complex mental process that depends on several brain networks working together over very short and longer time scales. One challenge in studying this process is that different brain imaging methods have different strengths: some show where activity happens better, while others show when it happens better. To get around this, the researchers combined functional MRI, which shows which brain areas are active, with EEG, which records the brain’s electrical activity, and used a mathematical tool called Continuous Wavelet Transform to examine changes in brain activity frequencies in the second after a word or sentence was presented. They compared natural language passages with scrambled words and found three brain network patterns that were more active during meaningful language processing. These included the main language network, a left-sided part of the default mode network, which is a set of brain regions often involved in internally directed thought, and another default mode subnetwork in both sides of the brain. Each network had its own “frequency fingerprint”: the language network was linked to longer-lasting theta activity along with bursts of beta and gamma activity, the first default mode network showed beta and gamma bursts, and the second default mode network was dominated by alpha activity. These patterns also related to language ability: differences in the language network’s frequency pattern were associated with how well people remembered what they read or heard, and reading comprehension depended partly on how strongly the language network and the alpha-dominant default mode network worked together. Overall, the study suggests that brain networks involved in language have distinct patterns of electrical activity that change over time and may help explain differences in language skill.

Fig. 1.��Stimuli presentation and fused fMRI-EEG frequency analysis. (A) Presentation of expository passages and non-sequential word baseline during both fMRI and EEG acquisition. (B) Fused fMRI-EEG analysis on subject-level inputs including passage (Pass) and word baseline (WB) to generate independent fused source components with subject weight loadings. (C) Continuous wavelet transform analysis on the EEG joint components to characterize frequency power over time throughout the post-stimulus window.

This study looked at how to make very high-resolution 7T diffusion MRI scans more accurate and reliable. Diffusion MRI is used to study the brain’s white matter pathways, but at such high field strength the images can be distorted, which can reduce anatomical accuracy and make results less reproducible. The researchers tested different ways of collecting and correcting the scans by using five healthy adults who each had two MRI sessions. They compared several methods, ranging from uncorrected scans to more advanced approaches that used multiple phase-encoding directions, which are different ways of collecting the image data to help correct distortion. They then checked how well each method lined up with standard anatomical MRI images and how consistent the measurements were when the scan was repeated. All of the correction methods improved image accuracy compared with uncorrected scans, but using a full set of reversed phase-encoding images worked better than the common approach of using only one reversed reference image. The best results came from using a four-direction phase-encoding scheme, which produced the most accurate images and the most reproducible measurements. This approach also allowed the researchers to reconstruct both large and very fine white matter pathways with high quality. Overall, the study shows that collecting diffusion MRI data in multiple directions is important for getting dependable results from high-resolution 7T brain scans.

Fig. 1.��Methodology. The highly oversampled acquisition (top) enabled creation of subset combinations of nine time-equivalent 10��min acquisitions (bottom). In the schematic, short blocks denote b��=��0 volumes and long blocks denote diffusion-weighted volumes (DWIs). Color encodes the phase-encoding axis (blue��=��AP/PA; red��=��LR/RL), and the shading direction within each block indicates phase-encoding polarity (e.g., AP vs PA; LR vs RL). A full 10��min acquisition includes 64 uniformly distributed diffusion weighted directions (with b��=��0 images interspersed every 16 volumes). This is repeated once for each of the four PE directions (AP, PA, RL, and LR). The nine corrected acquisitions depicted here fall into four categories: (1–4) single reverse PE b��=��0 (PA, AP, RL, LR); (5–6) full blip-up/blip-down with unique DWIs (AP-PA, RL-LR); (7–8) full blip-up/blip-down with repeated DWIs (AP-PAr, RL-LRr); (9) -way PE (AP-PA-RL-LR). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

]]> /valiant/2026/05/26/quantitative-imaging-of-iron-dysregulation-in-multiple-system-atrophy/ Tue, 26 May 2026 20:59:08 +0000 /valiant/?p=6780 Trujillo, Paula.; Hett, Kilian.; Cooper, Amy.; Brown, Amy E.; Donahue, Manus J.; McKnight, Colin D.; Bradbury, Margaret.; Wong, Cynthia.; Stamler, David.; Claassen, Daniel O. (2026).��.��NeuroImage, 334, 121965.��

Multiple system atrophy, or MSA, is a fast-moving brain disease that can be hard to diagnose early and monitor over time. This study looked at whether a special kind of MRI scan called quantitative susceptibility mapping, or QSM, can detect abnormal iron buildup in the brain, since iron imbalance may play a role in MSA and could be useful for tracking the disease. The researchers scanned 38 people with MSA, including 10 with early-stage disease who were followed again after 12 months, along with 43 people with Parkinson’s disease and 23 healthy adults of similar age. They measured iron-related changes in several brain regions, including the substantia nigra, globus pallidus, putamen, and dentate nucleus. Compared with both Parkinson’s disease and healthy controls, people with MSA had higher iron-related signal changes in the globus pallidus and substantia nigra, with smaller changes in the putamen. A more sensitive analysis that focused on higher values within each region detected these differences better than simple median measurements, suggesting it was better at picking up small, localized areas of iron buildup. Higher iron levels in the globus pallidus were also linked to worse clinical symptoms. In the 12-month follow-up, iron-related changes increased in the substantia nigra and globus pallidus, showing that these abnormalities can worsen over time. Overall, the findings suggest that QSM MRI may be a useful way to help diagnose MSA earlier, follow disease progression, and evaluate treatments aimed at reducing iron-related damage.

Fig. 1.��Representative QSM images from each diagnostic group.Quantitative susceptibility maps (QSM) for representative female participants matched for age across cohorts: a healthy control (HC, 67 years), Parkinson’s disease (PD, 62 years), PD from the bioMUSE study (69 years), multiple system atrophy (MSA) from bioMUSE (65 years), and MSA from the cross-sectional cohort (72 years). The PD (bioMUSE) participant was initially enrolled as MSA but later reclassified as PD based on longitudinal clinical evaluation. The MSA (bioMUSE) participant represents an early-stage case, whereas the MSA (cross-sectional) participant had more advanced disease. Each row displays a different subcortical region: the putamen (PT) and globus pallidus (GP) (top), the substantia nigra (SN) (middle), and the dentate nucleus (DN) (bottom). The leftmost column shows the QSM overlaid on the T1-weighted image to provide anatomical context. The second column shows the zoomed QSM for the HC participant with atlas-derived ROI outlines in red and anatomical labels. The remaining columns show the zoomed QSM for each diagnostic group. White arrows indicate the structures of interest in each row. Increases in magnetic susceptibility (brighter signal) are visible in the PT, GP, and SN in MSA participants compared with HC and PD. Susceptibility values are displayed in parts per million (ppm) using a grayscale colormap windowed from −0.1 to 0.2 ppm.

This study focuses on improving how passive resonators—devices used in MRI scanners to shape and strengthen radiofrequency (RF) fields—are designed and optimized. Normally, designing these structures requires��full-wave electromagnetic (EM) simulations, which model how RF fields behave in detail. While accurate, these simulations are extremely slow and computationally expensive, especially when many design variables (like different capacitor or inductor values) need to be tested.

To solve this problem, the researchers developed a faster method called a��co-simulation framework, which combines a single detailed EM simulation with simpler circuit-level calculations. In this approach, parts of the resonator are replaced with connection points (“ports”) during the initial simulation, allowing many different electrical configurations to be tested afterward without repeating the costly EM computation. They also integrated a��genetic algorithm��(a search method inspired by natural selection) to automatically explore thousands of design options and find the best configuration for enhancing RF fields in a specific target area.

The method was tested in several scenarios, from simple models to a realistic human head model, and produced results nearly identical to full EM simulations (with less than 1% error). Importantly, the optimization process took less than five minutes, compared to what would normally require extremely long computation times. Overall, this approach offers a much faster and scalable way to design passive MRI components, making it easier to improve image quality without the heavy computational cost of traditional methods.

Fig. 1.��Schematic diagram of the co-simulation principle. Incorporate the optimization stage, indicate the starting point of the method and reorganized the layout.

��

This study introduces a new wearable device—designed like a pair of glasses—that improves the quality of MRI scans of the eyes. MRI image quality is often described using the��signal-to-noise ratio (SNR), which compares the useful signal (clear image information) to background noise; higher SNR means clearer, more detailed images. Imaging the eyes is particularly challenging, especially at very high magnetic field strengths (such as 7 Tesla), where maintaining good image quality across both the eyes and the brain can be difficult.

The researchers created lightweight, 3D-printed “resonator glasses” that contain small electronic components called��LC loop resonators��(circuits that can enhance MRI signal locally). These resonators work wirelessly by interacting with the existing MRI head coil, meaning no modifications to the scanner hardware are needed. The team tested the device in both lab setups (phantoms, which simulate human tissue) and real human scans. They found that the glasses significantly improved image clarity in the eye region—boosting SNR by up to three times—while not reducing image quality in the rest of the brain.

Overall, this device offers a simple, low-cost way to enhance eye imaging during MRI scans without interfering with standard brain imaging. This could make it easier to study eye conditions or perform combined eye–brain imaging in clinical and research settings.

FIGURE 1

Circuit diagram (A) and CAD design (B) of the wireless resonator glasses.

]]> /valiant/2026/03/26/global-cortical-arousal-effects-in-fmri-reveal-brain-markers-of-state-and-trait-anxiety/ Thu, 26 Mar 2026 19:30:32 +0000 /valiant/?p=6334 Kimberly Kundert-Obando; Terra Lee; Caroline G. Martin; Kamalpreet Kaur; Juan Gomez Lagandara; Yamin Li; Jeffrey M. Harding; Shiyu Wang; Richard Song; Ruoqi Yang; Rithwik Guntaka; Sarah E. Goodale; Roza G. Bayrak; Lucina Q. Uddin; Martin Walter; Jeremy Hogeveen; Catie Chang (2026).��.��Cerebral Cortex, 36(2), bhag008.��

This study explores how brain activity measured with functional MRI (fMRI) can help better understand and personalize the diagnosis and treatment of anxiety. Anxiety is not just a psychological experience—it also involves physical responses in the body, such as changes in heart rate and alertness (called��arousal). These bodily and brain-wide states can influence fMRI signals across the entire brain, often referred to as “global” signals. Traditionally, these global signals have been treated as noise or interference, but the researchers investigated whether they might actually contain meaningful information about anxiety.

To do this, the team analyzed fMRI data to identify patterns related to both��autonomic physiological activity��(automatic body functions like heart rate) and��cortical arousal��(how alert or activated the brain is). They then examined how these patterns relate to two types of anxiety:��state anxiety��(temporary, situation-based anxiety) and��trait anxiety��(a person’s general tendency to feel anxious). The results showed clear links between these global brain signals and both forms of anxiety, with certain brain regions showing stronger associations. These patterns overlapped with well-known brain networks, including the��default mode network, which is involved in self-reflection and internal thoughts.

Overall, the findings suggest that these global fMRI signals carry useful information about how anxiety is represented in the brain. This insight could help improve how anxiety is measured and understood, potentially leading to more personalized approaches to diagnosis and treatment.

Fig 1. Spatial association between global components and anxiety. a and b) Areas in which the FAI was significantly associated with state and trait

anxiety. c and d) Areas in which the GS was significantly associated with state and trait anxiety (GS was analyzed using a negative contrast). Maps show

the t-statistics thresholded at P <0.05 corrected.

This study looks at how measurements of the spinal cord—such as its��cross-sectional area��(the size of the cord when viewed in a slice)—can be used as important indicators (biomarkers) for diagnosing and tracking neurological diseases like multiple sclerosis or spinal cord compression. Modern artificial intelligence methods can automatically identify and outline (segment) the spinal cord in MRI scans, but it is unclear whether these measurements stay consistent as models are updated with new data over time. This consistency is especially important when building “normal” reference values from healthy individuals.

To address this, the researchers developed a spinal cord segmentation model trained on a large and diverse dataset collected from 75 sites and over 1,600 participants, covering different MRI types and various spinal cord conditions. They also created a “lifelong learning” system that continuously monitors changes in measurements (called��morphometric drift) whenever the model is updated. This system automatically runs through a workflow (via GitHub Actions, an automated coding tool) to track how measurements evolve over time.

The results showed that the new model performs very well, accurately identifying the spinal cord even in challenging cases such as severe compression or tissue damage, with a high Dice score (a measure of how closely the model’s segmentation matches the true anatomy) of 0.95. The monitoring system also proved useful for quickly detecting any changes in measurements between model versions. Importantly, the study found that updates to the model caused only minimal shifts in spinal cord measurements, meaning the results remain stable and reliable. This allowed the researchers to safely update an existing database of normal spinal cord measurements. Overall, this work provides a reliable and transparent way to maintain consistency in AI-based medical measurements as models evolve.

Fig 1

Overview of the dataset and image characteristics. Representative axial slices of nine contrasts and the total of images used for each contrast in brackets, the orientation (axial/sagittal) along with the median resolution of images. The respective doughnut chart illustrates the proportion of clinical status among the scanned participants, including healthy controls (HC), patients with radiologically isolated syndrome (RIS), patients with multiple sclerosis (MS), and their different phenotypes, including primary progressive (PPMS) and relapsing-remitting (RRMS), patients with amyotrophic lateral sclerosis (ALS), patients with neuromyelitis optica spectrum disorder (NMOSD), pre-decompression acute traumatic SCI (AcuteSCI), post-decompression traumatic spinal cord injury (SCI), degenerative cervical myelopathy (DCM), and syringomyelia (SYR; not shown). Labels indicate the phenotype associated with the patient, with their respective colors shared across contrast sets.

This study followed cognitively normal adults over time to determine which magnetic resonance imaging (MRI) biomarkers best predict future cognitive impairment. Researchers examined 154 different MRI-based measurements in 509 participants from the Baltimore Longitudinal Study of Aging who were age 50 or older and cognitively normal at the start of the study. Participants underwent repeated cognitive testing and 3 Tesla (3T) MRI scans, including T1- and T2-weighted imaging to assess brain structure and diffusion tensor imaging (DTI) to measure white matter microstructural integrity. The analyses accounted for factors such as age and other confounders and also examined differences by sex and amyloid beta (Aβ) status, a biological marker associated with Alzheimer’s disease.

Over an average follow-up of 4.6 years, individuals who later developed cognitive impairment showed greater declines in white matter integrity compared to those who remained cognitively stable. These changes were especially pronounced in major white matter tracts, including the corpus callosum, cingulum bundle, and inferior fronto-occipital fasciculus, which are pathways that connect different brain regions. To a lesser extent, thinning and atrophy in the temporal lobe were also linked to later impairment. The associations between brain changes and future cognitive decline were stronger in men and in individuals who were amyloid-positive.

Overall, the findings suggest that early changes in white matter microstructure, as measured by DTI, are particularly sensitive indicators of future mild cognitive impairment (MCI) and dementia. Certain MRI metrics may therefore be especially useful for identifying risk in people who are still cognitively normal.

��

FIGURE 1

Study overview. Participants were selected from the BLSA neuroimaging substudy based on cognitively normal (CN) status and age 50 or older at baseline. The study data included longitudinal cognitive assessments, clinical diagnoses (Dx), 3T magnetic resonance imaging scans, and baseline plasma biomarkers related to Alzheimer’s disease and related dementias, specifically amyloid beta 42/40, collected between 2008 and 2019. The subsequently impaired (SI) group (also CN at baseline) included individuals who later developed mild cognitive impairment (MCI) or dementia or were “Impaired, not MCI/dementia.” Impairment onset dates ranged from 2012 to 2019 (≈1- to 9-year interval).