High-quality Magnetic Resonance Imaging Research Articles

High-quality multimodal MRI with simultaneous EEG using conductive ink and polymer-thick film nets

Objective. Combining magnetic resonance imaging (MRI) and electroencephalography (EEG) provides a powerful tool for investigating brain function at varying spatial and temporal scales. Simultaneous acquisition of both modalities can provide unique information that a single modality alone cannot reveal. However, current simultaneous EEG-fMRI studies are limited to a small set of MRI sequences due to the image quality and safety limitations of commercially available MR-conditional EEG nets. We tested whether the Inknet2, a high-resistance polymer thick film based EEG net that uses conductive ink, could enable the acquisition of a variety of MR image modalities with minimal artifacts by reducing the radiofrequency-shielding caused by traditional MR-conditional nets.Approach. We first performed simulations to model the effect of the EEG nets on the magnetic field and image quality. We then performed phantom scans to test image quality with a conventional copper EEG net, with the new Inknet2, and without any EEG net. Finally, we scanned five human subjects at 3 Tesla (3 T) and three human subjects at 7 Tesla (7 T) with and without the Inknet2 to assess structural and functional MRI image quality.Main results. Across these simulations, phantom scans, and human studies, the Inknet2 induced fewer artifacts than the conventional net and produced image quality similar to scans with no net present.Significance. Our results demonstrate that high-quality structural and functional multimodal imaging across a variety of MRI pulse sequences at both 3 T and 7 T is achievable with an EEG net made with conductive ink and polymer thick film technology.

An open-access lumbosacral spine MRI dataset with enhanced spinal nerve root structure resolution

Spinal cord injury (SCI) profoundly affects an individual’s ability to move. Fortunately, recent advancements in neuromodulation, particularly the spatio-temporal epidural electrical stimulation (EES) targeting the spinal nerve roots, promoted rapid rehabilitation of SCI patients. Such neuromodulation techniques require precise anatomical modelling of spinal cord. However, the lack of spine imaging datasets, especially high-quality magnetic resonance imaging (MRI) datasets highlighting nerve roots, hinders the translation of EES into medical practice. To address this problem, we introduce an open-access lumbosacral spine MRI dataset acquired in 14 healthy adults, using constructive interference in steady state (CISS) sequence, double echo steady state (DESS) sequence, and T2-weight turbo spin echo (T2-TSE) sequence, with enhanced nerve root resolution. The dataset also includes the corresponding anatomical annotations of nerve roots and the final reconstructed 3D spinal cord models. The quality of our dataset is assessed using image quality metrics implemented in MRI quality control tool (MRIQC). Our dataset provides a valuable platform to promote a wide range of spinal cord neuromodulation research and collaboration among neurorehabilitation engineers.

Meniscal Lesions in Multi-Ligament Knee Injuries.

Multi-ligament knee injuries (MLKIs) are rare and complex knee lesions and are potentially associated with intra-articular injuries, especially meniscal tears. Understanding the meniscal tear patterns involved in MLKI can help the orthopedic surgeon treat these complex injuries. The purpose of this study was to describe the incidence, classification, and treatment of meniscal injuries in a cohort of patients with MLKIs and carry out an updated review of the evidence available. Descriptive retrospective study. Patients with a history of reconstructive surgery for MLKI performed between 2013 and 2023 were included. Informed consent was obtained from all patients included in the study. Patient demographics, magnetic resonance imaging (MRI) study, and operative reports were reviewed. Groups were then formed based on ligament injury patterns. Meniscal tears were identified by MRI and through diagnostic arthroscopy for each patient. The association between meniscal lesions and injury patterns was calculated through Fisher's exact test. Agreement between the presence of meniscal tear on MRI and in diagnostic arthroscopy was measured using the kappa test. The sensitivity and specificity of MRI were calculated. We inferred the presence of a meniscal tear by injury pattern using the Agresti-Coull confidence interval. For the statistical analysis, a significance of 5% and a confidence interval of 95% were considered. Seventy patients with MLKIs were included, with a mean age of 30.69years (SD 10.65). Forty-seven patients had meniscal lesions (67.1%). Of them, 6 had only medial meniscus tears, 31 had only lateral meniscus tears, and 10 had lesions of both menisci, comprising 57 meniscal lesions in total. An anterior cruciate ligament (ACL) + medial collateral ligament/posteromedial corner (MCL/PMC) was the most common injury pattern (52.86% of all patients). Of these 37 patients, 78.38% had meniscal injuries, and most of them (68.97%) were only lateral meniscus injuries. The odds ratio (OR) of having a meniscal tear when having an ACL + medial-side injury was 4.83 (95% CI; 0.89-26.17). Patients with ACL + lateral-side injury pattern had meniscal tears in 42.86%. The lateral meniscus was involved in 100% of these patients. 62.5% of medial meniscus injuries were treated by meniscal repair, and 37.5% by partial meniscectomy. 58.54% of lateral meniscus injuries were treated by meniscal repair, and 39.02% by partial meniscectomy. Agreement calculated using the kappa test between MRI and diagnostic arthroscopy for medial meniscal lesions was 78.57%, and for lateral meniscal lesions was 84.29%. The ligament injury pattern and the side of the injured collateral ligament influenced the incidence and laterality of meniscal damage. ACL + medial-side injuries were shown to have significantly greater meniscal damage compared to other injury patterns.It is crucial to have a high index of suspicion, obtain a high-quality MRI, and arthroscopically evaluate any possible meniscal lesions in MLKIs.

Perspectives on technology: Prostate Imaging-Reporting and Data System (PI-RADS) interobserver variability.

To explore the topic of Prostate Imaging-Reporting and Data System (PI-RADS) interobserver variability, including a discussion of major sources, mitigation approaches, and future directions. A narrative review of PI-RADS interobserver variability. PI-RADS was developed in 2012 to set technical standards for prostate magnetic resonance imaging (MRI), reduce interobserver variability at interpretation, and improve diagnostic accuracy in the MRI-directed diagnostic pathway for detection of clinically significant prostate cancer. While PI-RADS has been validated in selected research cohorts with prostate cancer imaging experts, subsequent prospective studies in routine clinical practice demonstrate wide variability in diagnostic performance. Radiologist and biopsy operator experience are the most important contributing drivers of high-quality care among multiple interrelated factors including variability in MRI hardware and technique, image quality, and population and patient-specific factors such as prostate cancer disease prevalence. Iterative improvements in PI-RADS have helped flatten the curve for novice readers and reduce variability. Innovations in image quality reporting, administrative and organisational workflows, and artificial intelligence hold promise in improving variability even further. Continued research into PI-RADS is needed to facilitate benchmark creation, reader certification, and independent accreditation, which are systems-level interventions needed to uphold and maintain high-quality prostate MRI across entire populations.

Adaptive Knowledge Distillation for High-Quality Unsupervised MRI Reconstruction With Model-Driven Priors.

Magnetic Resonance Imaging (MRI) reconstruction has made significant progress with the introduction of Deep Learning (DL) technology combined with Compressed Sensing (CS). However, most existing methods require large fully sampled training datasets to supervise the training process, which may be unavailable in many applications. Current unsupervised models also show limitations in performance or speed and may face unaligned distributions during testing. This paper proposes an unsupervised method to train competitive reconstruction models that can generate high-quality samples in an end-to-end style. Firstly teacher models are trained by filling the re-undersampled images and compared with the undersampled images in a self-supervised manner. The teacher models are then distilled to train another cascade model that can leverage the entire undersampled k-space during its training and testing. Additionally, we propose an adaptive distillation method to re-weight the samples based on the variance of teachers, which represents the confidence of the reconstruction results, to improve the quality of distillation. Experimental results on multiple datasets demonstrate that our method significantly accelerates the inference process while preserving or even improving the performance compared to the teacher model. In our tests, the distilled models show 5%-10% improvements in PSNR and SSIM compared with no distillation and are 10 times faster than the teacher.

DFUSNN: zero-shot dual-domain fusion unsupervised neural network for parallel MRI reconstruction

Objective. Recently, deep learning models have been used to reconstruct parallel magnetic resonance (MR) images from undersampled k-space data. However, most existing approaches depend on large databases of fully sampled MR data for training, which can be challenging or sometimes infeasible to acquire in certain scenarios. The goal is to develop an effective alternative for improved reconstruction quality that does not rely on external training datasets. Approach. We introduce a novel zero-shot dual-domain fusion unsupervised neural network (DFUSNN) for parallel MR imaging reconstruction without any external training datasets. We employ the Noise2Noise (N2N) network for the reconstruction in the k-space domain, integrate phase and coil sensitivity smoothness priors into the k-space N2N network, and use an early stopping criterion to prevent overfitting. Additionally, we propose a dual-domain fusion method based on Bayesian optimization to enhance reconstruction quality efficiently. Results. Simulation experiments conducted on three datasets with different undersampling patterns showed that the DFUSNN outperforms all other competing unsupervised methods and the one-shot Hankel-k-space generative model (HKGM). The DFUSNN also achieves comparable results to the supervised Deep-SLR method. Significance. The novel DFUSNN model offers a viable solution for reconstructing high-quality MR images without the need for external training datasets, thereby overcoming a major hurdle in scenarios where acquiring fully sampled MR data is difficult.

Safe and Informed Use of Gadolinium-Based Contrast Agent in Body Magnetic Resonance Imaging: Where We Were and Where We Are.

Gadolinium-based contrast agents (GBCAs) have helped to improve the role of magnetic resonance imaging (MRI) for the diagnosis and treatment of diseases. There are currently nine different commercially available gadolinium-based contrast agents (GBCAs) that can be used for body MRI cases, and which are classifiable according to their structures (cyclic or linear) or biodistribution (extracellular-space agents, target/specific-agents, and blood-pool agents). The aim of this review is to illustrate the commercially available MRI contrast agents, their effect on imaging, and adverse reaction on the body, with the goal to lead to their proper selection in different clinical contexts. When we have to choose between the different GBCAs, we have to consider several factors: (1) safety and clinical impact; (2) biodistribution and diagnostic application; (3) higher relaxivity and better lesion detection; (4) higher stability and lower tissue deposit; (5) gadolinium dose/concentration and lower volume injection; (6) pulse sequences and protocol optimization; (7) higher contrast-to-noise ratio at 3.0 T than at 1.5 T. Knowing the patient's clinical information, the relevant GBCAs properties and their effect on body MRI sequences are the key features to perform efficient and high-quality MRI examination.

Neuroimaging Findings in US Government Personnel and Their Family Members Involved in Anomalous Health Incidents

US government personnel stationed internationally have reported anomalous health incidents (AHIs), with some individuals experiencing persistent debilitating symptoms. To assess the potential presence of magnetic resonance imaging (MRI)-detectable brain lesions in participants with AHIs, with respect to a well-matched control group. This exploratory study was conducted at the National Institutes of Health (NIH) Clinical Center and the NIH MRI Research Facility between June 2018 and November 2022. Eighty-one participants with AHIs and 48 age- and sex-matched control participants, 29 of whom had similar employment as the AHI group, were assessed with clinical, volumetric, and functional MRI. A high-quality diffusion MRI scan and a second volumetric scan were also acquired during a different session. The structural MRI acquisition protocol was optimized to achieve high reproducibility. Forty-nine participants with AHIs had at least 1 additional imaging session approximately 6 to 12 months from the first visit. AHIs. Group-level quantitative metrics obtained from multiple modalities: (1) volumetric measurement, voxel-wise and region of interest (ROI)-wise; (2) diffusion MRI-derived metrics, voxel-wise and ROI-wise; and (3) ROI-wise within-network resting-state functional connectivity using functional MRI. Exploratory data analyses used both standard, nonparametric tests and bayesian multilevel modeling. Among the 81 participants with AHIs, the mean (SD) age was 42 (9) years and 49% were female; among the 48 control participants, the mean (SD) age was 43 (11) years and 42% were female. Imaging scans were performed as early as 14 days after experiencing AHIs with a median delay period of 80 (IQR, 36-544) days. After adjustment for multiple comparisons, no significant differences between participants with AHIs and control participants were found for any MRI modality. At an unadjusted threshold (P < .05), compared with control participants, participants with AHIs had lower intranetwork connectivity in the salience networks, a larger corpus callosum, and diffusion MRI differences in the corpus callosum, superior longitudinal fasciculus, cingulum, inferior cerebellar peduncle, and amygdala. The structural MRI measurements were highly reproducible (median coefficient of variation <1% across all global volumetric ROIs and <1.5% for all white matter ROIs for diffusion metrics). Even individuals with large differences from control participants exhibited stable longitudinal results (typically, <±1% across visits), suggesting the absence of evolving lesions. The relationships between the imaging and clinical variables were weak (median Spearman ρ = 0.10). The study did not replicate the results of a previously published investigation of AHIs. In this exploratory neuroimaging study, there were no significant differences in imaging measures of brain structure or function between individuals reporting AHIs and matched control participants after adjustment for multiple comparisons.

Role of visceral fat on postoperative complications and relapse in patients with Crohn's disease after ileocecal resection: Is it overrated?

The role of visceral fat in disease development, particularly in Crohn´s disease (CD), is significant. However, its preoperative prognostic value for postoperative complications and CD relapse after ileocecal resection (ICR) remains unknown. This study aims to assess the predictive potential of preoperatively measured visceral and subcutaneous fat in postoperative complications and CD recurrence using magnetic resonance imaging (MRI). The primary endpoint was postoperative anastomotic leakage of the ileocolonic anastomosis, with secondary endpoints evaluating postoperative complications according to the Clavien Dindo classification and CD recurrence at the anastomosis. We conducted a retrospective analysis of 347 CD patients who underwent ICR at our tertiary referral center between 2010 and 2020. We included 223 patients with high-quality preoperative MRI scans, recording demographics, postoperative outcomes, and CD recurrence rates at the anastomosis. To assess adipose tissue distribution, we measured total fat area (TFA), visceral fat area (VFA), subcutaneous fat area (SFA), and abdominal circumference (AC) at the lumbar 3 (L3) level using MRI cross-sectional images. Ratios of these values were calculated. None of the radiological variables showed an association with anastomotic leakage (TFA p = 0.932, VFA p = 0.982, SFA p = 0.951, SFA/TFA p = 0.422, VFA/TFA p = 0.422), postoperative complications, or CD recurrence (TFA p = 0.264, VFA p = 0.916, SFA p = 0.103, SFA/TFA p = 0.059, VFA/TFA p = 0.059). Radiological visceral obesity variables were associated with postoperative outcomes or clinical recurrence in CD patients undergoing ICR. Preoperative measurement of visceral fat measurement is not specific for predicting postoperative complications or CD relapse.

Current MR imaging of cartilage in the context of knee osteoarthritis (part2) : Cartilage pathologies and their assessment

High-quality magnetic resonance (MR) imaging is essential for the precise assessment of the knee joint and plays a key role in the diagnostics, treatment and prognosis. Intact cartilage tissue is characterized by asmooth surface, uniform tissue thickness and an organized zonal structure, which are manifested as depth-dependent signal intensity variations. Cartilage pathologies are identifiable through alterations in signal intensity and morphology and should be communicated based on a precise terminology. Cartilage pathologies can show hyperintense and hypointense signal alterations. Cartilage defects are assessed based on their depth and should be described in terms of their location and extent. The following symptom constellations are of overarching clinical relevance in image reading and interpretation: symptom constellations associated with rapidly progressive forms of joint degeneration and unfavorable prognosis, accompanying symptom constellations mostly in connection with destabilizing meniscal lesions and subchondral insufficiency fractures (accelerated osteoarthritis) as well as symptoms beyond the "typical" degeneration, especially when adiscrepancy is observed between (minor) structural changes and (major) synovitis and effusion (inflammatory arthropathy).

Structural neuroimaging changes associated with subjective cognitive decline from a clinical sample

BackgroundAlzheimer’s disease (AD) is characterized by progressive deterioration of cognitive functions. Some individuals with subjective cognitive decline (SCD) are in the early phase of the disease and subsequently progress through the AD continuum. Although neuroimaging biomarkers could be used for the accurate and early diagnosis of preclinical AD, the findings in SCD samples have been heterogeneous. This study established the morphological differences in brain magnetic resonance imaging (MRI) findings between individuals with SCD and those without cognitive impairment based on a clinical sample of patients defined according to SCD-Initiative recommendations. Moreover, we investigated baseline structural changes in the brains of participants who remained stable or progressed to mild cognitive impairment or dementia. MethodsThis study included 309 participants with SCD and 43 healthy controls (HCs) with high-quality brain MRI at baseline. Among the 99 subjects in the SCD group who were followed clinically, 32 progressed (SCDp) and 67 remained stable (SCDnp). A voxel-wise statistical comparison of gray and white matter (WM) volume was performed between the HC and SCD groups and between the HC, SCDp, and SCDnp groups. XTRACT ATLAS was used to define the anatomical location of WM tract damage. Region-of-interest (ROI) analyses were performed to determine brain volumetric differences. White matter lesion (WML) burden was established in each group. ResultsVoxel-based morphometry (VBM) analysis revealed that the SCD group exhibited gray matter atrophy in the middle frontal gyri, superior orbital gyri, superior frontal gyri, right rectal gyrus, whole occipital lobule, and both thalami and precunei. Meanwhile, ROI analysis revealed decreased volume in the left rectal gyrus, bilateral medial orbital gyri, middle frontal gyri, superior frontal gyri, calcarine fissure, and left thalamus. The SCDp group exhibited greater hippocampal atrophy (p < 0.001) than the SCDnp and HC groups on ROI analyses. On VBM analysis, however, the SCDp group exhibited increased hippocampal atrophy only when compared to the SCDnp group (p < 0.001). The SCD group demonstrated lower WM volume in the uncinate fasciculus, cingulum, inferior fronto-occipital fasciculus, anterior thalamic radiation, and callosum forceps than the HC group. However, no significant differences in WML number (p = 0.345) or volume (p = 0.156) were observed between the SCD and HC groups. ConclusionsThe SCD group showed brain atrophy mainly in the frontal and occipital lobes. However, only the SCDp group demonstrated atrophy in the medial temporal lobe at baseline. Structural damage in the brain regions was anatomically connected, which may contribute to early memory decline.

Testing a Home Solution for Preparing Young Children for an Awake MRI: A Promising Smartphone Application.

Thanks to its non-invasive nature and high-resolution imaging capabilities, magnetic resonance imaging (MRI) is a valuable diagnostic tool for pediatric patients. However, the fear and anxiety experienced by young children during MRI scans often result in suboptimal image quality and the need for sedation/anesthesia. This study aimed to evaluate the effect of a smartphone application called COSMO@home to prepare children for MRI scans to reduce the need for sedation or general anesthesia. The COSMO@home app was developed incorporating mini-games and an engaging storyline to prepare children for learning goals related to the MRI procedure. A multicenter study was conducted involving four hospitals in Belgium. Eligible children aged 4-10 years were prepared with the COSMO@home app at home. Baseline, pre-scan, and post-scan questionnaires measured anxiety evolution in two age groups (4-6 years and 7-10 years). Eighty-two children participated in the study, with 95% obtaining high-quality MRI images. The app was well-received by children and parents, with minimal technical difficulties reported. In the 4-6-year-old group (N = 33), there was a significant difference between baseline and pre-scan parent-reported anxiety scores, indicating an increase in anxiety levels prior to the scan. In the 7-10-year-old group (N = 49), no significant differences were observed between baseline and pre-scan parent-reported anxiety scores. Overall, the COSMO@home app proved to be useful in preparing children for MRI scans, with high satisfaction rates and successful image outcomes across different hospitals. The app, combined with minimal face-to-face guidance on the day of the scan, showed the potential to replace or assist traditional face-to-face training methods. This innovative approach has the potential to reduce the need for sedation or general anesthesia during pediatric MRI scans and its associated risks and improve patient experience.

Brain Injury Outcomes after Adjuvant Erythropoietin Neuroprotection for Moderate or Severe Neonatal Hypoxic-Ischemic Encephalopathy: A Report from the HEAL Trial

Introduction: Erythropoietin (Epo) is a putative neuroprotective therapy that did not improve overall outcomes in a phase 3 randomized controlled trial for neonates with moderate or severe hypoxic-ischemic encephalopathy (HIE). However, HIE is a heterogeneous disorder, and it remains to be determined whether Epo had beneficial effects on a subset of perinatal brain injuries. Methods: This study was a secondary analysis of neuroimaging data from the High-dose Erythropoietin for Asphyxia and Encephalopathy (HEAL) Trial, which was conducted from 2016 to 2021 at 17 sites involving 23 US academic medical centers. Participants were neonates >36 weeks’ gestation undergoing therapeutic hypothermia for moderate or severe HIE who received 5 doses of study drug (Epoetin alpha 1,000 U/kg/dose) or placebo in the first week of life. Treatment assignment was stratified by trial site and severity of encephalopathy. The primary outcome was the locus, pattern, and acuity of brain injury as determined by three independent readers using a validated HIE Magnetic Resonance Imaging (MRI) scoring system. Results: Of the 500 infants enrolled in HEAL, 470 (94%) had high quality MRI data obtained at a median of 4.9 days of age (IQR: 4.5–5.8). The incidence of injury to the deep gray nuclei, cortex, white matter, brainstem and cerebellum was similar between Epo and placebo groups. Likewise, the distribution of injury patterns was similar between groups. Among infants imaged at less than 8 days (n = 414), 94 (23%) evidenced only acute, 93 (22%) only subacute and 89 (21%) both acute and subacute injuries, with similar distribution across treatment groups. Conclusion: Adjuvant erythropoietin did not reduce the incidence of regional brain injury. Subacute brain injury was more common than previously reported, which has key implications for the development of adjuvant neuroprotective therapies for this population.

Chiral Iron Oxide Supraparticles Enable Enantiomer-Dependent Tumor-Targeted Magnetic Resonance Imaging.

The chemical, physical and biological effects of chiral nanomaterials have inspired general interest and demonstrated important advantages in fundamental science. Here, chiral iron oxide supraparticles (Fe3 O4 SPs) modified by chiral penicillamine (Pen) molecules with g-factor of ≈2 × 10-3 at 415nm are fabricated, and these SPs act as high-quality magnetic resonance imaging (MRI) contrast agents. Therein, the transverse relaxation efficiency and T2 -MRI results demonstrated chiral Fe3 O4 SPs have a r2 relaxivity of 157.39 ± 2.34 mM-1 ·S-1 for D-Fe3 O4 SPs and 136.21 ± 1.26 mM-1 ·S-1 for L-Fe3 O4 SPs due to enhanced electronic transition dipole moment for D-Fe3 O4 SPs compared with L-Fe3 O4 SPs. The in vivo MRI results show that D-Fe3 O4 SPs exhibit two-fold lower contrast ratio than L-Fe3 O4 SPs, which enhances targeted enrichment in tumor tissue, such as prostate cancer, melanoma and brain glioma tumors. Notably, it is found that D-Fe3 O4 SPs have 7.7-fold higher affinity for the tumor cell surface receptor cluster-of-differentiation 47 (CD47) than L-Fe3 O4 SPs. These findings uncover that chiral Fe3 O4 SPs act as a highly effective MRI contrast agent for targeting and imaging broad tumors, thus accelerating the practical application of chiral nanomaterials and deepening the understanding of chirality in biological and non-biological environments.

Deep learning method for magnetic resonance imaging fluid-attenuated inversion recovery image synthesis

Magnetic resonance imaging(MRI) can obtain multi-modal images with different contrast, which provides rich information for clinical diagnosis. However, some contrast images are not scanned or the quality of the acquired images cannot meet the diagnostic requirements due to the difficulty of patient's cooperation or the limitation of scanning conditions. Image synthesis techniques have become a method to compensate for such image deficiencies. In recent years, deep learning has been widely used in the field of MRI synthesis. In this paper, a synthesis network based on multi-modal fusion is proposed, which firstly uses a feature encoder to encode the features of multiple unimodal images separately, and then fuses the features of different modal images through a feature fusion module, and finally generates the target modal image. The similarity measure between the target image and the predicted image in the network is improved by introducing a dynamic weighted combined loss function based on the spatial domain and K-space domain. After experimental validation and quantitative comparison, the multi-modal fusion deep learning network proposed in this paper can effectively synthesize high-quality MRI fluid-attenuated inversion recovery (FLAIR) images. In summary, the method proposed in this paper can reduce MRI scanning time of the patient, as well as solve the clinical problem of missing FLAIR images or image quality that is difficult to meet diagnostic requirements.

Robust Unsupervised Super-Resolution of Infant MRI via Dual-Modal Deep Image Prior.

Magnetic resonance imaging (MRI) is commonly used for studying infant brain development. However, due to the lengthy image acquisition time and limited subject compliance, high-quality infant MRI can be challenging. Without imposing additional burden on image acquisition, image super-resolution (SR) can be used to enhance image quality post-acquisition. Most SR techniques are supervised and trained on multiple aligned low-resolution (LR) and high-resolution (HR) image pairs, which in practice are not usually available. Unlike supervised approaches, Deep Image Prior (DIP) can be employed for unsupervised single-image SR, utilizing solely the input LR image for de novo optimization to produce an HR image. However, determining when to stop early in DIP training is non-trivial and presents a challenge to fully automating the SR process. To address this issue, we constrain the low-frequency k-space of the SR image to be similar to that of the LR image. We further improve performance by designing a dual-modal framework that leverages shared anatomical information between T1-weighted and T2-weighted images. We evaluated our model, dual-modal DIP (dmDIP), on infant MRI data acquired from birth to one year of age, demonstrating that enhanced image quality can be obtained with substantially reduced sensitivity to early stopping.

2D brain MRI image synthesis based on lightweight denoising diffusion probabilistic model

In recent years, brain health has received increasing attention, but conventional acquisition of brain MRI (magnetic resonance imaging) images still suffer from issues such as missing data, artifacts, and high costs, which hinders research and diagnosis. With the application of deep learning in medical image synthesis, low-cost, efficient, and high-quality medical MRI synthesis techniques have become a prominent research focus and have gradually matured. However, traditional methods for synthesizing magnetic resonance imaging (MRI) mostly rely on generative adversarial networks, which require fine-tuning of parameters and learning rates to achieve stringent Nash equilibrium conditions, leading to problems such as gradient explosions and mode collapse. Building upon the latest research in synthetic models DDPM (denoising diffusion probabilistic model), we propose a novel approach for 2D brain MRI image synthesis based on a lightweight denoising diffusion probabilistic model. This method improves the attention module in the denoising diffusion probabilistic model to make it more lightweight. Additionally, we adopt the smooth L1 loss function as a replacement for the traditional mean absolute error (L1 loss) by comparing the error between the 2D brain MRI images with added noise and the real noise for training the model. Finally, we validate the proposed model on the MRI Brain Tumor Classification dataset, demonstrating that it achieves high-quality synthesis results while significantly reducing the parameter count of the DDPM model.

Management of a challenging dura-embedded anterior inferior cerebellar artery loop during a retrosigmoid hearing-preserving vestibular schwannoma resection: microsurgical technique and operative video. Illustrative case.

Anatomical variants of the anterior inferior cerebellar artery (AICA), such as an anomalous "AICA loop" embedded in the dura and bone of the subarcuate fossa, increase the complexity and risk of vestibular schwannoma resections. Classically, osseous penetrating AICA loops are the most challenging to mobilize, as the dura must be dissected and the surrounding petrous bone must be drilled to mobilize the AICA away from the surgical corridor and out of harm. The authors present a rare case of a dura-embedded, osseous-penetrating AICA loop encountered during a hearing-preserving retrosigmoid approach in which they demonstrate safe and efficient microdissection and mobilization of the AICA loop without having to drill the surrounding bone. Although preoperative recognition of potentially dangerous AICA loops has been challenging, thin-sliced petrous bone computed tomography scanning and high-quality magnetic resonance imaging can be useful in preoperative diagnosis. Furthermore, this report suggests that a retrosigmoid approach is superior, as it allows early intradural recognition and proximal vascular control and facilitates more versatile mobilization of AICA loops.

Accelerating image reconstruction for multi-contrast MRI based on Y-Net3.

As one of the significant preoperative imaging modalities in medical diagnosis, Magnetic resonance imaging (MRI) takes a long scanning time due to its special imaging principle. We propose an innovative MRI reconstruction strategy and data consistency method based on deep learning to reconstruct high-quality brain MRIs from down-sampled data and accelerate the MR imaging process. Sixteen healthy subjects undergoing T1-weighted spin-echo (SE) and T2-weighted fast spin-echo (FSE) sequences by a 1.5T MRI scanner were recruited. A Y-Net3+ network was used to facilitate the high-quality MRI reconstruction through context information. In addition, the existing data consistency fidelity method was improved. The difference between the reconstructed K-space and the original K-space was shorten by the linear regression algorithm. Therefore, the redundant artifacts derived from under-sampling were avoided. The Structural Similarity (SSIM) and Peak Signal to Noise Ratio (PSNR) were applied to quantitatively evaluate image reconstruction performance of different down-sampling patterns. Compared with the classical Y-Net, Y-Net3+ network improved SSIM and PSNR of MRI images from 0.9164±0.0178 and 33.2216±3.2919 to 0.9387±0.0363 and 35.1785±3.3105, respectively, under compressed sensing reconstruction with acceleration factor of 4. The improved network increases signal-to-noise ratio and adds more image texture information in the reconstructed images. Furthermore, in the process of data consistency, linear regression analysis was used to reduce the difference between the reconstructed K-space and the original K-space, so that the SSIM and PSNR were increased to 0.9808±0.0081 and 40.9254±1.1911, respectively. The improved Y-Net combined with data consistency fidelity method elucidates its potential in reconstructing high-quality T2-weighted images from the down-sampled data by fully exploring the T1-weighted information. With the advantage of avoiding down-sampled artifacts, the improved network exhibits remarkable clinical promise for fast MRI applications.

MHAN: Multi-Stage Hybrid Attention Network for MRI reconstruction and super-resolution

High-quality magnetic resonance imaging (MRI) affords clear body tissue structure for reliable diagnosing. However, there is a principal problem of the trade-off between acquisition speed and image quality. Image reconstruction and super-resolution are crucial techniques to solve these problems. In the main field of MR image restoration, most researchers mainly focus on only one of these aspects, namely reconstruction or super-resolution. In this paper, we propose an efficient model called Multi-Stage Hybrid Attention Network (MHAN) that performs the multi-task of recovering high-resolution (HR) MR images from low-resolution (LR) under-sampled measurements. Our model is highlighted by three major modules: (i) an Amplified Spatial Attention Block (ASAB) capable of enhancing the differences in spatial information, (ii) a Self-Attention Block with a Data-Consistency Layer (DC-SAB), which can improve the accuracy of the extracted feature information, (iii) an Adaptive Local Residual Attention Block (ALRAB) that focuses on both spatial and channel information. MHAN employs an encoder–decoder architecture to deeply extract contextual information and a pipeline to provide spatial accuracy. Compared with the recent multi-task model T2Net, our MHAN improves by 2.759 dB in PSNR and 0.026 in SSIM with scaling factor ×2 and acceleration factor 4× on T2 modality.