共找到 20 条结果
Non-invasive Magnetic Resonance Imaging on clinical MR-scanners at spatial resolution ≂1mm3 could in principle be used also for insights in the structure of valuable ancient objects like books if the spatial resolution could be improved. We demonstrate, that Magnetic-Resonance-Microscopy is even able to visualize printed letters at thickness < 30 µm on superposed paper-sheets. The physical-technical methodology is based on prototype hardware installed as insert on a human MR-scanner relying on strong magnetic and gradient-fields with sensitive radiofrequency-sensors. A negative contrast mechanism, adding MR-visible chemically-inert liquid, is necessary, its removal being potentially harmful to valuable paper sheets. For visualization of text firstly the letter thickness (≂20 µm) and exact positions were determined before adjusting optimized resolution and field-of-view (FOV) in measurement and reprocessing of 3D-data for slice-positioning in the plane of the paper sheets. The advantages of a semi-automatic data processing method for visualization on the bowed paper-sheets are demonstrated. In contrast to micro-Computed tomography based on absorption contrast, MR-microscopy can visualize non-metallic printed letters and offers higher spatial resolution than Terahertz imaging. As a preclinical MR-imaging tool (with limited FOV) it is more widely available than neutron-tomography. The reported MRM-technology might be also of interest for radiological high-resolution applications such as MR-based histology.
Although Magnetic Resonance Imaging (MRI) is gaining momentum in forensic and post-mortem settings, its exact role in forensic investigations is yet to be determined. The current review aims to chart current and potential roles of MRI in forensic investigations, providing an overview of existing pertinent scientific literature. A mapping review was conducted in accordance with the PRISMA guidelines. Medline, Web of Science, Embase and the Cochrane Library were searched using database-specific syntaxes. Eligible articles included reviews, original research, case reports, letters, laboratory studies, and dissertations written in English. Articles not mentioning MRI, involving only functional MRI, and non-forensic clinical post-mortem examinations were excluded. Articles were categorised using thematic analysis. Data extraction included first-author country of origin, year of publication, level of evidence, and mention of MRI-protocol. The initial search yielded 16,184 papers, 525 of which were included. Main focus of forensic MRI research is in medicolegal examination (n = 285), identification (n = 158), and process optimisation (n = 155); 73 articles were multi-categorical, and myriad subcategories were identified (foetus/child/adult, living/postmortem, anatomical regions, practice/guidelines/infrastructure, etc.). Articles were published over a 34-year period, originating from 34 countries. There were 235 level C-, 289 level B-, and 2 level A-evidence articles. 268 articles mentioned MRI protocols. The extensive amount of research across myriad subcategories highlights vast potential of MRI in forensic investigation. However, the overall level of evidence at this time is low, lacking standardisation. Further organization, standardization, and high-quality research are needed to clarify how best to apply MRI in forensic settings.
Magnetic resonance imaging (MRI) is a powerful tool for diagnosing and monitoring brain diseases, but its low sensitivity can hinder early detection. To address this challenge, we utilized chemical exchange saturation transfer (CEST) MRI, which greatly enhances sensitivity for detecting low-concentration compounds. In this study, we developed a CEST contrast agent based on a recombinant adeno-associated viruses (rAAVs) encoding the protamine-1 (PRM1) MRI reporter gene. CEST MRI revealed that PRM1 contrast agent effectively highlighted caudate putamen region after injection of the rAAVs into the mouse brain, clearly distinguishing it from the surrounding tissue, with no observable damage. This method provides a sensitive, metal-free CEST contrast agent for in vivo brain cell detection, demonstrating potential for both diagnostic and therapeutic applications in brain diseases.
After the fabrication of magnetic resonance superconducting magnets, the magnetic field inhomogeneity needs to be accurately measured for subsequent shimming. However, conventional measurement methods are susceptible to magnetic fields, have poor compatibility, and are difficult to adapt to various types of magnets. This paper proposes a new field measuring system based on a three-axis movable platform. The system utilizes non-magnetic materials and an innovative hand-wheel lifting design that can be adapted to various aperture magnets, thus obviating the necessity for electrically driven equipment and addressing safety concerns in strong magnetic fields. In addition, the measurement system offers high accuracy up to 1 mm and a wide measurable range. The fields of 3 T and 7 T magnets were mapped using the designed system with diameter of spherical volume (DSV) of 160 mm and 130 mm, respectively. Experimental results demonstrate that the magnetic field measurement system has strong compatibility and can accurately map the magnetic field at arbitrary positions, which is critical for shimming studies.
Mirror invariance is the cognitive tendency to perceive mirror-image objects as identical. Mirrored letters, however, are distinct orthographic units and must be identified as different despite having the same shape. Consistent with this phenomenon, a small, localized region in the ventral visual stream, the Visual Word Form Area (VWFA), exhibits repetition suppression to both identical and mirror pairs of objects but only to identical, not mirror, pairs of letters ( Pegado et al., 2011), a phenomenon named mirror invariance "breaking". The ability of congenitally blind individuals to "break" mirror invariance for pairs of mirrored Braille letters has been demonstrated behaviorally ( de Heering et al., 2018, Korczyk et al., 2024). However, its neural underpinnings have not yet been investigated. Here, in an fMRI repetition suppression paradigm, congenitally blind individuals (8 males and 10 females) recognized pairs of everyday objects and Braille letters in identical ("p" and "p"), mirror ("p" and "q"), and different ("p" and "z") orientations. We found repetition suppression for identical and mirror pairs of everyday objects in the parietal and ventral-lateral occipital cortex, indicating that mirror-invariant object recognition engages the ventral visual stream in tactile modality as well. However, repetition suppression for identical but not mirrored pairs of Braille letters was found not in the VWFA, but in broad areas of the left parietal cortex and the lateral occipital cortex. These results suggest that reading-related orthographic processes in blind individuals depend on different neural computations than those of the sighted.
A single spin-echo, acquired in the presence of a constant magnetic field gradient, has a phase proportional to the average velocity of the sample and an amplitude determined by the distribution of velocities for laminar flow in a pipe. If we make assumptions about symmetry, it is therefore possible to reconstruct a velocity profile from a series of spin echoes acquired with different echo times, TE = 2 τ . The velocity profile encompasses a range of shear stresses and so describes the shear-rate dependence of the fluid viscosity, making this a rapid, non-invasive rheometric measurement. The appeal of this approach lies in its simplicity: a constant, uniform magnetic field gradient can be readily constructed using low-field permanent magnets and the resulting magnetic resonance instrument is relatively cheap and easily sited. In order to extend the range of fluids to which such a rheometer can be applied, we have designed a pre-measurement polarization unit to maximize the polarization of fluids with a long T 1 , such as aqueous solutions.
Low back pain (LBP) is the leading cause of disability worldwide. Despite guidelines advising against routine imaging, a significant proportion of patients are referred for lumbar spine magnetic resonance imaging (MRI) without an appropriate indication. This study aims to assess the impact of the intervention on the appropriateness of referrals and the proportion of referred patients from general practitioners (GPs) who underwent lumbar MRI. A quasi-experimental design with pre- and post-intervention periods was employed. We included GP referrals for lumbar spine MRI at Silkeborg Regional Hospital, Denmark, among adults ≥ 18 years from 1 January 2019 to 31 August 2023. The intervention, which consisted of template-based correspondence letters and a guidance booklet, was implemented on 1 September 2022. Referrals were classified as either 'appropriate' or 'inappropriate' according to international guidelines. Interrupted time-series analysis with segmented regression was applied to evaluate change over time in referral appropriateness and the proportion of lumbar MRIs within 90 days of referral. A total of 5222 referrals were received (80.6% pre and 19.4% post intervention). Patient characteristics were comparable across the periods. The proportion of appropriate referrals increased from 49.7 to 60.0%. At the onset of the intervention, the level of appropriate referrals increased by 6.50% points (95% CI 0.71, 12.29), and the trend changed by 0.98 (95% CI 0.36, 1.59), reaching 0.90 (95% CI 0.30, 1.49). This implied a continued increase in the pre-post difference. The proportion of referred patients having lumbar MRIs decreased from 92.2 to 75.7%. At intervention onset, the level of lumbar MRIs changed by - 17.06 (95% CI - 23.25, - 10.88). However, post-intervention the trend changed, becoming positive (0.66, 95% CI - 0.05, 1.37), implying that the number of lumbar MRIs could return to pre-intervention level over time. The intervention increased the proportion of appropriate referrals for lumbar MRI. For the proportion of MRIs performed, we found an immediate reduction, but this may not be sustained over time.
Severe atherosclerotic internal carotid artery stenosis may progress to complete internal carotid artery occlusion (ICAO). Therefore, ICAO represents an advanced form of carotid artery disease. We sought to investigate the association between ICAO with atherosclerotic disease in other arterial beds and vascular risk factors and to identify the patient implications of the diagnosis of ICAO. Using the term "Internal carotid artery occlusion," a search of PubMed/MEDLINE, Scopus, and Embase between 1980 and 2025 revealed 10,588 results. After exclusion of case reports, letters to the Editor and Editorials, 5771 reports were identified. Following meticulous screening of the identified reports, 28 studies specifically addressing patient with ICAO cohorts were included in the final analysis. A quantitative and qualitative synthesis analysis was performed. A questionnaire was subsequently developed and sent out to 63 participants from the United States (n = 21) and several European countries (n = 42), aiming to achieve consensus regarding the optimal management of patients with ICAO. Three participants did not respond. The Consensus Coordinator abstained from voting to avoid introducing bias, resulting in a final voting panel of 60 participants. Across included studies, the proportion of patients with ICAO presenting with neurologic symptoms varied widely, ranging from 38% to 100%, whereas approximately 24% to 27% of patients were asymptomatic at the time of diagnosis. Consensus (≥75%) was achieved in 11 of the 17 (64.7%) prespecified statements. Most participants agreed that atherosclerotic ICAO represents a systemic manifestation of advanced atherosclerosis rather than isolated cerebrovascular pathology (56/60; 93.3%). Duplex ultrasound study should be used as the first-line diagnostic tool for suspected ICAO, with computed tomography angiography or magnetic resonance angiography confirmation if necessary (59/60; 98.3%). Optimal medical therapy (including antiplatelet, antihypertensives, statins, and glycemic control) remains the cornerstone of ICAO management (59/60; 98.3%). Lifestyle and metabolic risk factor optimization, smoking cessation, optimizing body weight, a healthy diet and exercise, should be strongly advised in all patients with ICAO (60/60; 100%). Most participants concurred that ICAO revascularization should be centralized in specialized vascular-neuro centers equipped for intraoperative neuromonitoring and advanced hemodynamic control (57/60; 95.0%). Finally, most participants agreed that current evidence for ICAO intervention is insufficient and that a global registry should be created to record outcomes and guide future trials (56/60; 93.3%). This international, multispecialty consensus highlights ICAO as a marker of advanced, systemic atherosclerosis. Management should emphasize comprehensive evaluation for multisystem vascular disease and aggressive modification of cardiovascular risk factors. Best medical therapy remains the cornerstone of the management of patients with ICAO, with conservative or invasive interventions considered selectively based on symptom status, anatomic considerations, procedural risk, and institutional expertise, to reduce the overall cardiovascular disease burden.
Major depressive disorder (MDD) is a highly prevalent psychiatric disorder marked by disrupted brain dynamics. However, the neural mechanisms underlying remission remain poorly understood, particularly regarding common neural markers across diverse therapeutic interventions. Emerging evidence suggests that temporal brain dynamics and their hierarchical organization, referred to as metastates, serve as sensitive markers of individual variability across cognitive functions. In this study, we evaluated whether metastate dynamics derived from resting-state functional magnetic resonance imaging (rs-fMRI) differed according to remission status across pharmacotherapy, psychotherapy, and neuromodulation. This multicenter observational study with 370 participants included 229 individuals with depression and 141 healthy control participants. The depression cohort comprised individuals undergoing cognitive behavioral therapy (n = 92), pharmacotherapy (n = 59), electroconvulsive therapy (n = 50), and repetitive transcranial magnetic stimulation (n = 28). rs-fMRI data were analyzed to derive metastate dynamics, and comparisons were made according to remission status across treatment modalities. Two distinct metastates were identified, one associated with higher-order cognitive brain regions and another linked to sensory and motor systems. Participants who achieved remission exhibited greater predictability in transitions between brain states within metastates, supporting higher-order cognitive functions. This altered transition pattern was accompanied by alterations in the anticorrelation between the default mode and executive function networks, which may underlie the increased predictability. Remission from MDD may involve a reorganization of hierarchical brain dynamics-particularly in systems supporting cognitive control-and may offer a potential treatment modality-independent biomarker of remission.
2-hydroxyglutarate (2-HG) is a key metabolic biomarker for identifying IDH-mutant gliomas. Non-invasive and accurate detection of 2-HG is of great significance for the early diagnosis of diseases and dynamic monitoring of therapeutic efficacy. However, conventional magnetic resonance spectroscopy (MRS) faces challenges in detecting 2-HG in vivo, mainly due to the overlap of its resonance peaks with those of metabolites such as glutamate (Glu) and N-acetylaspartate (NAA). Although the long echo time (TE) filtering method can separate signals to a certain extent, it is often accompanied by peak distortion and signal attenuation, which limits its clinical application. To address this problem, this study proposes a 2-HG-targeted detection sequence based on optimal control pulses. By applying optimal control pulses to regulate the state evolution of a 14-spin system composed of 2-HG, Glu, and NAA molecules, the study achieves selective retention of 2-HG signals and suppression of other molecular signals. In experimental verification conducted on phantoms and IDH-mutant glioma animal models, the targeted sequence exhibited excellent signal resolution performance: it efficiently retained 2-HG signals and achieved approximately 95% and 98% suppression of Glu and NAA signals, respectively. To further verify the quantitative reliability of the targeted sequence, the 2-HG concentrations measured by this sequence were compared with those obtained by liquid chromatography-tandem mass spectrometry (LC-MS/MS). A high linear correlation was found between the two sets of results, which fully confirms the accuracy of non-invasive quantitative detection of 2-HG using the targeted sequence.
We present a functional magnetic resonance (fMRI) dataset collected as part of an adversarial collaboration aimed at arbitrating between the Global Neuronal Workspace theory (GNWT) and the Integrated Information Theory (IIT) of consciousness. Participants (N = 118) were presented with suprathreshold visual stimuli belonging to four different categories (faces, objects, letters, false fonts) with three orientations (front, left, right view), and three durations (0.5, 1.0, 1.5 seconds). Participants were asked to identify infrequent targets that changed in each block, thereby rendering two categories task-relevant and two task-irrelevant. The simplicity of the experimental design and of the task given to the participants ensures that these data are broadly reusable. Besides testing predictions from other theories of consciousness, these data can be used to examine various aspects of visual processing. The anonymized data were converted to Brain Imaging Data Structure (BIDS), and can be easily accessed through a web platform or an API. The dataset contains quality reports, demographics, behavioral performance, and eye-tracking data. We also provide code for preprocessing and analyzing the data.
To address the need for observing the electrolyte influenced by dissolved metal ions during the charge and discharge process of lithium-ion battery, a fast-switching broadband radio frequency (RF) switch for broadband nuclear magnetic resonance (NMR) systems was designed. This design specifically aims to overcome the challenge where signals exhibiting fast relaxation-often containing critical information-are lost due to the long "dead time" of conventional NMR instruments. The switch utilizes a differential architecture paired with a single-ended drive circuit. Its innovative feature is the inclusion of common-mode chokes in the bias circuit, which suppresses switching spikes and accelerates the transition speed. Electrical testing results confirm its superior performance: within a 3 MHz-100 MHz bandwidth, the switch achieves an insertion loss of less than 0.86 dB and isolation greater than 85 dB, coupled with an input third-order intercept point (IIP3) linearity of 55.50 dBm. Crucially, the measured switching time is only 175.3 ns, and the switching spike is limited to 6.09 mV. Experimental validation with paramagnetic additives (Mn2+, Fe2+) demonstrated that the switch's short dead time preserves the signal-to-noise ratio (SNR) of short transverse relaxation time (T 2) signals while a 5 μs delay reduced the SNR by over 45%, underscoring the practical importance of fast switching. In practical NMR experiments using LiPF6 electrolyte, the switch successfully detected the free induction decay (FID) signals of 1H, 7Li, and 19F nuclei. The resultant short switching time minimizes the NMR front-end dead time, which is highly advantageous for detecting samples with short T 2 relaxation and broad spectra.
Deuterium metabolic imaging (DMI) enables noninvasive, spatially resolved mapping of in vivo metabolism by tracking deuterium-labeled substrates and downstream products, yet it is frequently signal-to-noise-ratio (SNR) limiteddue to the low gyromagnetic ratio of 2H and low metabolite concentration. Balanced steady-state free precession (bSSFP) can improve scan-time SNR efficiency for DMI, but operation in the short repetition time (TR)regime restricts readout duration and the number of acquired free induction decay (FID) points. The resulting loss of spectral resolvability increases metabolite overlap, renders separation ill-conditioned, and destabilizes fast Fourier transform (FFT)-basedprocessing and IDEAL-type model-based least-squares decomposition via strong cross-talk. We propose a physics-embedded CycleGAN for robust metabolite mapping in short-readout DMI using unpaired training. The key innovation is to embed an analytical bSSFP-DMI forward model directly into the cycle-consistency pathway by replacing one generator with a deterministic physics operator, thereby enforcing measurement consistency without requiring paired ground truth. A U-Net generator estimates metabolite maps and an off-resonance field map, with an off-resonance field map (B0) constrained to be spatially smooth via spline-basis parameterization. The physics operator synthesizes voxel wise time-domain signals, enabling a signal-domain cycle loss. Across simulation stress tests and in vivo mouse experiments, our method consistently suppresses inter-metabolite leakage and preserves spatial structure as readout length decreases, outperforming FFT peak integration and IDEAL-type fitting in the short-readout regime.
Nuclear magnetic resonance (NMR) spectroscopy is a powerful tool for analyzing molecular structure and composition. However, traditional NMR experiments suffer from long acquisition times, especially in multidimensional NMR spectroscopy. This problem, to some extent, limits broader applications of NMR techniques. Various methods have been proposed to accelerate sampling, including non-uniform sampling (NUS), multi-FID acquisition (MFA), Hadamard encoding, Fourier encoding, spatial encoding Ultrafast 2D NMR (UF2DNMR), and so on. The review focuses on rapid sampling methods developed in contemporary China, introducing their fundamental principles and applications while discussing their respective advantages and disadvantages.
Distributions of nuclear magnetic resonance (NMR) relaxation times provide detailed information about the water in wood. This study documents the water dynamics analysis of T 2 and T 1 distributions for saturated delignified sapwood (DSW), delignified heartwood (DHW) and lignocellulose (LC) samples at different temperatures. Results indicate that below the freezing point of bulk water, free water freezes, causing its signal to disappear from the distribution. Then, the low temperature distributions of the unfrozen bound water contain more information about its components, with DSW, DHW and LC containing two distinct states of bound water (OH bound water (B-water) and more freely bound water (C-water)). Furthermore, it was observed that within the temperature range of -3°C to -60°C, B-water in DSW, DHW and LC maintained a higher unfrozen water content (UWC) value than C-water, and the T 1/T 2 ratios for B-water were consistently higher than that for C-water, indicating that B-water has a greater antifreeze capacity. T 2 and T 1 distributions offer different kinds of information about water components, and all peaks within the distribution have been assigned.
Language has referential and emotive uses. Referential language depicts events and can be verified or falsified by comparison with the event, whereas emotive language lacks the events upon which one can determine the truth value of the sentence. Using functional magnetic resonance imaging (fMRI), we investigated the brain mechanisms of the processing of the two semantically distinct forms of language in a paradigm of empathy for pain conveyed through single short sentences. In the emotive condition, subjective expressions of pain, such as "I have a toothache", were used. In the referential condition, an objective description of events in which an individual would experience pain, such as "I cut my finger", were presented. Assuming each stimulus elicits activation proportional to pain intensity ratings, we combined referential and emotive conditions and performed an analysis with a parametric modulation model. It revealed activation in the anterior cingulate cortex and, slightly below the threshold, the anterior insula, that are involved in the perception of one's own pain. Without pooling the two conditions, only the referential condition yielded activation in the two regions. Crucially, although both emotive and referential language activated multiple regions, emotive sentences activated the right temporoparietal junction, whereas referential sentences activated the precuneus/retrosplenial cortex, parahippocampal cortex, and the language network in the left hemisphere. These results suggest that a complementary interhemispheric network supports the processing of both types of language.
Subjective cognitive decline (SCD) is a preclinical stage of mild cognitive impairment (MCI). Although dance training has been shown to be beneficial for mental health, cognitive function, and neural activity in older adults with MCI, its effect on SCD remains unclear. This study aimed to examine the effects of dance training on the aforementioned factors and on oxytocin secretion in older adults with SCD. Participants (aged 65-84 years) were assigned to either the intervention group (n = 22) with a 12-week dance training program or the control group without any alternative training (n = 22). Apathy, depression, Montreal Cognitive Assessment scores, urinary oxytocin levels, and resting-state functional magnetic resonance imaging indices, including amplitude of low-frequency fluctuations (ALFF) and functional connectivity (FC), were evaluated pre- and post-intervention. Compared to the control group, the intervention group exhibited significantly higher urinary oxytocin levels and significantly higher ALFF in the left medial orbitofrontal cortex post-intervention. Moreover, the intervention group showed more enhanced FC between the left medial orbitofrontal cortex and the left precuneus post-intervention than the control group. However, mental health or cognitive performance was not significantly different between the groups. Our results are particularly important in light of previous findings that older adults with SCD show a reduced FC between the medial orbitofrontal cortex and the precuneus, and that oxytocin levels are positively associated with the prefrontal-amygdala oxytocinergic circuit in socioemotional processing. Thus, dance training may contribute to socioemotional resilience-related neural and molecular adaptations in SCD.
Deep neural networks (DNNs) trained on magnetic resonance images can estimate global brain age (GBA), which reflects women's neurological disease risk. GBA gap (GBAG), the difference between GBA and chronological age (CA), quantifies excessive global aging; local BAG (LBAG) has not been examined despite allowing voxelwise resolution. Using a novel DNN architecture, we estimate LBAG for 12,284 UK Biobank females with chronological ages (CAs) ranging between 46 and 82 years (y) and quantify how it relates to cognition and women's health variables (CA at menopause, reproductive span, menopausal hormone therapy (HT), contraceptive use (CU), number of births). Women with longer reproductive spans (-0.042/y ≤ β ≤ -0.037/y, p < 0.001) had older CAs at menopause onset (-0.052/y ≤ β ≤ -0.046/y, p < 0.001), more births (-0.230 ≤ β ≤ -0.190 per birth, p < 0.001) and younger brains (more negative LBAGs, younger GBAs); a 1-unit increase in each of these variables reflects an LBAG change of β y. Left temporal lobe effects of CA at menopause onset are strongest (-0.0517 ≤ β ≤ -0.0510, p < 0.001). Cognitive scores are related to LBAGs negatively and most strongly in subcortical and right-hemisphere cortex (-0.021 ≤ β ≤ -0.017 per score unit, p < 0.01). In postmenopausal women, delayed regional brain aging is predicted by longer endogenous hormone exposure indexed by later menopause onset, longer reproductive span, and more births. This research highlights the complex role of women's health factors upon brain aging and related cognitive trajectories.
This review summarizes current insights into Reversible cerebral vasoconstriction syndrome (RCVS) diagnosis, management, and outcomes. RCVS is a cerebrovascular disorder characterized by recurrent thunderclap headaches and transient segmental vasoconstriction of cerebral arteries, typically resolving within 3 months. A comprehensive database search was performed across MEDLINE (via PubMed), Embase, Scopus, and the Cochrane Library. Although often self-limiting, RCVS may cause complications such as subarachnoid hemorrhage, ischemic stroke, and cerebral edema. Triggers include vasoactive substances, pregnancy, postpartum state, and physical or emotional stress. Differentiating RCVS from conditions like primary angiitis of the central nervous system, aneurysmal subarachnoid hemorrhage, and cerebral venous thrombosis is essential because clinical and imaging features may overlap, whereas treatments differ. Advances in neuroimaging, especially magnetic resonance angiography and vessel wall imaging, have enhanced diagnostic accuracy. Management focuses on eliminating triggers and symptomatic support. Calcium channel blockers are frequently used, although their impact on disease evolution remains uncertain. Although most patients recover without major sequelae, chronic symptoms such as long-term headaches and neuropsychological symptoms, including cognitive impairment, underscore the need for ongoing follow-up and suggest a post-RCVS syndrome. Reversible cerebral vasoconstriction syndrome (RCVS) causes sudden, severe “thunderclap” headaches due to temporary narrowing of brain arteries that typically resolves within 3 months, although serious complications like stroke can occur. This review examines the historical understanding of RCVS, its common triggers (including stress, physical exertion, certain medications, and pregnancy), and current approaches to diagnosis using advanced brain imaging techniques. A sudden, explosive headache requires immediate medical attention, and even after recovery, some patients experience ongoing headaches or cognitive changes that may indicate a “post‐RCVS” syndrome.
Accurate segmentation of brain tumors from magnetic resonance imaging (MRI) is a critical task in neuro-oncology, directly impacting diagnosis, treatment planning, and patient outcomes. This study presents a hybrid deep learning model that integrates the VGG19 convolutional neural network as encoder and the U-Net architecture as decoder for automated tumor segmentation. The model is trained and validated using the lower-grade glioma (LGG) segmentation dataset from the cancer imaging archive (TCIA), which includes annotated FLAIR MRI scans of LGG cases. By leveraging the deep feature extraction capabilities of VGG19 and the spatial precision of U-Net, the proposed model achieves superior segmentation accuracy while maintaining architectural simplicity. Quantitative evaluation demonstrates a Dice of 92.20%, intersection over union (IoU) of 85.53%, and area under the curve (AUC) of 95.90%, outperforming several state-of-the-art models in the literature. This framework offers a reliable and efficient solution for clinical MRI analysis and has strong potential for integration into computer-aided diagnostic systems. The results indicate that hybrid convolutional neural network (CNN) architectures can significantly enhance the accuracy and robustness of brain tumor segmentation tasks. The novelty of this work lies in the seamless integration of VGG19's pre-trained deep features with U-Net's skip connections, providing a lightweight yet highly accurate model that reduces computational complexity compared to multi-branch or attention-based architectures, while achieving superior performance on the LGG dataset.