MR Imaging Research Articles

Ligand-Induced Atomically Segregation-Tunable Alloy Nanoprobes for Enhanced Magnetic Resonance Imaging.

Bimetallic iron-noble metal alloy nanoparticles have emerged as promising contrast agents for magnetic resonance imaging (MRI) due to their biocompatibility and facile control over the element distribution. However, the inherent surface energy discrepancy between iron and noble metal often leads to Fe atom segregation within the nanoparticle, resulting in limited iron-water molecule interactions and, consequently, diminished relaxometric performance. In this study, we present the development of a class of ligand-induced atomically segregation-tunable alloy nanoprobes (STAN) composed of bimetallic iron-gold nanoparticles. By manipulating the oxidation state of Fe on the particle surface through varying molar ratios of oleic acid and oleylamine ligands, we successfully achieve surface Fe enrichment. Under the application of a 9 T MRI system, the optimized STAN formulation, characterized by a surface Fe content of 60.1 at %, exhibits an impressive r1 value of 2.28 mM-1·s-1, along with a low r2/r1 ratio of 6.2. This exceptional performance allows for the clear visualization of hepatic tumors as small as 0.7 mm in diameter in vivo, highlighting the immense potential of STAN as a next-generation contrast agent for highly sensitive MR imaging.

Current Practice: Rationale for Screening Children with Hereditary Hemorrhagic Telangiectasia for Brain Vascular Malformations.

Hereditary hemorrhagic telangiectasia is an autosomal dominant vascular dysplasia characterized by mucocutaneous telangiectasias, recurrent epistaxis, and organ vascular malformations including in the brain, which occur in about 10% of patients. These brain vascular malformations include high-flow AVMs and AVFs as well as low-flow capillary malformations. High-flow lesions can rupture, causing neurologic morbidity and mortality. International guidelines for the diagnosis and management of hereditary hemorrhagic telangiectasia recommend screening children for brain vascular malformations with contrast enhanced MR imaging at hereditary hemorrhagic telangiectasia diagnosis. Screening has not been uniformly adopted by some practitioners who contend that screening is not justified. Arguments against screening include application of short-term data from the adult A Randomized Trial of Unruptured Brain Arteriovenous Malformations (ARUBA) trial of unruptured sporadic brain AVMs to children with hereditary hemorrhagic telangiectasia as well as concerns about administration of sedation or IV contrast and causing patients or families increased anxiety. In this article, a multidisciplinary group of experts on hereditary hemorrhagic telangiectasia reviewed data that support screening guidelines and counter arguments against screening. Children with hereditary hemorrhagic telangiectasia have a preponderance of high-flow lesions including AVFs, which have the highest rupture risk. The rupture risk among children is estimated at about 0.7% per lesion per year and is additive across lesions and during a lifetime. ARUBA, an adult clinical trial of expectant medical management versus treatment of unruptured brain AVMs, favored medical management at 5 years but is not applicable to pediatric patients with hereditary hemorrhagic telangiectasia given the life expectancy of a child. Additionally, interventional, radiosurgical, and surgical techniques have improved with time. Experienced neurovascular experts can prospectively determine the best treatment for each child on the basis of local resources. The "watch and wait" approach to imaging means that children with brain vascular malformations will not be identified until a potentially life-threatening and deficit-producing intracerebral hemorrhage occurs. This expert group does not deem this to be an acceptable trade-off.

Multiparametric Characterization and Spatial Distribution of Different MS Lesion Phenotypes.

MS lesions exhibit varying degrees of axonal and myelin damage. A comprehensive description of lesion phenotypes could contribute to an improved radiologic evaluation of smoldering inflammation and remyelination processes. This study aimed to identify in vivo distinct MS lesion types using quantitative susceptibility mapping and susceptibility mapping-weighted imaging and to characterize them through T1-relaxometry, myelin mapping, and diffusion MR imaging. The spatial distribution of lesion phenotypes in relation to ventricular CSF was investigated. MS lesions of 53 individuals were categorized into iso- or hypointense lesions, hyperintense lesions, and paramagnetic rim lesions, on the basis of their appearance on quantitative susceptibility mapping alone, according to published criteria, and with the additional support of susceptibility mapping-weighted imaging. Susceptibility values, T1-relaxation times, myelin and free water fractions, intracellular volume fraction, and the orientation dispersion index were compared among lesion phenotypes. The distance of the geometric center of each lesion from the ventricular CSF was calculated. Eight hundred ninety-six MS lesions underwent the categorization process using quantitative susceptibility mapping and susceptibility mapping-weighted imaging. The novel use of susceptibility mapping-weighted images, which revealed additional microvasculature details, led us to re-allocate several lesions to different categories, resulting in a 35.6% decrease in the number of paramagnetic rim lesions, a 22.5% decrease in hyperintense lesions, and a 17.2% increase in iso- or hypointense lesions, with respect to the categorization based on quantitative susceptibility mapping only. The outcome of the categorization based on the joint use of quantitative susceptibility mapping and susceptibility mapping-weighted imaging was that 44.4% of lesions were iso- or hypointense lesions, 47.9% were hyperintense lesions, and 7.7% were paramagnetic rim lesions. A worsening gradient was observed from iso- or hypointense lesions to hyperintense lesions to paramagnetic rim lesions in T1-relaxation times, myelin water fraction, free water faction, and intracellular volume fraction. Paramagnetic rim lesions were located closer to ventricular CSF than iso- or hypointense lesions. The volume of hyperintense lesions was associated with a more severe disease course. Quantitative susceptibility mapping and susceptibility mapping-weighted imaging allow in vivo classification of MS lesions into different phenotypes, characterized by different levels of axonal and myelin loss and spatial distribution. Hyperintense lesions and paramagnetic rim lesions, which have the most severe microstructural damage, were more often observed in the periventricular WM and were associated with a more severe disease course.

Imaging findings and MRI patterns in a cohort of 18q chromosomal abnormalities.

The abnormalities of long arm of chromosome 18 (18q) constitute a complex spectrum. We aimed to systematically analyze their MRI features. We hypothesized that there would be variable but recognizable white matter and structural patterns in this cohort. In this retrospective cohort study, we included pediatric patients with a proven abnormality of 18q between 2000-2022. An age and sex matched control cohort was also constructed. Thirty-six cases, median MRI age 19.6 months (4.3 - 59.3), satisfied our inclusion criteria. Majority were females (25, 69%, F:M ratio 2.2:1). Fifty MR imaging studies were analyzed and 35 (70%) had delayed myelination. Two independent readers scored brain myelination with excellent interrater reliability. Three recognizable evolving MRI patterns with distinct age distributions and improving myelination scores were identified - PMD-like (9.9 months, 37), intermediate (22 months, 48) and washed-out pattern (113.6 months, 53). Etiologically, MRIs were analyzed across three subgroups - 18q- (34, 69%), trisomy 18 (10, 21%) and ring chromosome 18 (5, 10%). Ring chromosome 18 had the highest myelination lag (27, P value = 0.005) and multifocal white matter changes (P value = 0.001). Trisomy 18 had smaller pons and cerebellar dimensions (APD pons P value = 0.002, CC vermis P value <0.001 and TCD P value = 0.04). In this cohort of 18q chromosomal abnormalities, MRI revealed recognizable patterns correlating with improving brain myelination. Imaging findings appear to be on a continuum with more severe white matter abnormalities in ring chromosome 18 and greater prevalence of structural abnormalities of pons and cerebellum in trisomy 18. 18q-: 18q deletion; CC: corpus callosum; CC-APD: CC anteroposterior diameter; FOD: fronto-occipital diameter; TCD: transverse cerebellar diameter; APD: anteroposterior diameter; CCD: craniocaudal diameter; MBP: myelin basic protein; PMD: Pelizaeus-Merzbacher Disease; GWMD: gray-white matter differentiation.

Gaining Insight into Updated MR Imaging for Quantitative Assessment of Cartilage Injury in Knee Osteoarthritis.

Knee Osteoarthritis (KOA) entails progressive cartilage degradation, reviewed via MRI for morphology, biochemical composition, and microtissue alterations, discussing clinical advantages, limitations, and research applicability. Compositional MRI, like T2/T2* mapping, T1rho mapping, gagCEST, dGEMRIC, sodium imaging, diffusion-weighted imaging, and diffusion-tensor imaging, provide insights into cartilage injury in KOA. These methods quantitatively measure collagen, glycosaminoglycans, and water content, revealing important information about biochemical compositional and microstructural alterations. Innovative techniques like hybrid multi-dimensional MRI and diffusion-relaxation correlation spectrum imaging show potential in depicting initial cartilage changes at a sub-voxel level. Integration of automated image analysis tools addressed limitations in manual cartilage segmentation, ensuring robust and reproducible assessments of KOA cartilage. Compositional MRI techniques reveal microstructural changes in cartilage. Multi-dimensional MR imaging assesses biochemical alterations in KOA-afflicted cartilage, aiding early degeneration identification. Integrating artificial intelligence enhances cartilage analysis, optimal diagnostic accuracy for early KOA detection and monitoring.

Deep-learning-based motion correction using multichannel MRI data: a study using simulated artifacts in the fastMRI dataset.

Deep learning presents a generalizable solution for motion correction requiring no pulse sequence modifications or additional hardware, but previous networks have all been applied to coil-combined data. Multichannel MRI data provide a degree of spatial encoding that may be useful for motion correction. We hypothesize that incorporating deep learning for motion correction prior to coil combination will improve results. A conditional generative adversarial network was trained using simulated rigid motion artifacts in brain images acquired at multiple sites with multiple contrasts (not limited to healthy subjects). We compared the performance of deep-learning-based motion correction on individual channel images (single-channel model) with that performed after coil combination (channel-combined model). We also investigate simultaneous motion correction of all channel data from an image volume (multichannel model). The single-channel model significantly (p < 0.0001) improved mean absolute error, with an average 50.9% improvement compared with the uncorrected images. This was significantly (p < 0.0001) better than the 36.3% improvement achieved by the channel-combined model (conventional approach). The multichannel model provided no significant improvement in quantitative measures of image quality compared with the uncorrected images. Results were independent of the presence of pathology, and generalizable to a new center unseen during training. Performing motion correction on single-channel images prior to coil combination provided an improvement in performance compared with conventional deep-learning-based motion correction. Improved deep learning methods for retrospective correction of motion-affected MR images could reduce the need for repeat scans if applied in a clinical setting.

BRAIN TUMOR PERCEPTION USING DEEP LEARNING

The human brain serves as the central component of the human body, orchestrating its functionality. Brain tumors emerge from abnormal cell growth and division within the brain, potentially leading to the development of brain cancer. Computer vision plays a pivotal role in healthcare by alleviating the burden of making precise decisions. Among imaging modalities like magnetic resonance imaging (MRI), computed tomography (CT) scans, and X-rays, MRI scans stand out as one of the most prevalent and safest methods for obtaining detailed images, capable of detecting minute abnormalities. Our research endeavors to explore the diverse applications of brain MRI in detecting brain cancer. In this study, we employed the bilateral filter (BF) to eliminate noise from MR images before processing. Subsequently, the tumor region was delineated using binary thresholding and Convolutional Neural Network (CNN) segmentation techniques. The datasets were divided into training, testing, and validation subsets, enabling our machine to discern the presence of brain tumors in subjects. Various performance metrics including accuracy, sensitivity, and specificity were utilized to evaluate the outcomes. Dense layers within the CNN architecture efficiently extract features from brain MRI images, enhancing diagnostic accuracy. The experiment was conducted primarily on MRI data due to its ability to provide comprehensive insights into cellular

Advanced MRI techniques in abdominal imaging.

Magnetic resonance imaging (MRI) is a crucial modality for abdominal imaging evaluation of focal lesions and tissue properties. However, several obstacles, such as prolonged scan times, limitations in patients' breath-hold capacity, and contrast agent-associated artifacts, remain in abdominal MR images. Recent techniques, including parallel imaging, three-dimensional acquisition, compressed sensing, and deep learning, have been developed to reduce the scan time while ensuring acceptable image quality or to achieve higher resolution without extending the scan duration. Quantitative measurements using MRI techniques enable the noninvasive evaluation of specific materials. A comprehensive understanding of these advanced techniques is essential for accurate interpretation of MRI sequences. Herein, we therefore review advanced abdominal MRI techniques.

Comparative analysis of intraoperative and imaging features of invasive growth in pituitary adenomas.

Most pituitary adenomas (PAs), also termed pituitary neuroendocrine tumors, are benign in nature and can be treated effectively by surgical resection, medical treatment, and in special cases by radiotherapy. However, invasive growth can be an important feature of a more aggressive behavior and adverse prognosis. The extension of PAs into the cavernous sinus can be categorized according to the Knosp criteria on magnetic resonance imaging (MRI). Comparative analyses of MRI features and intraoperative findings of invasive growth regarding different clinical factors are still scarce. We performed a retrospective single-center analysis of 764 PAs that were surgically treated between October 2004 and April 2018. Invasive growth was assessed according to the surgical reports and preoperative MRI (Knosp criteria). Clinical data, such as patient age at diagnosis and gender, histopathological adenoma type, and extent of resection, were collected. Invasive features on MRI were seen in 24.4% (Knosp grades 3A-4, 186/764) of the cases. Intraoperatively, invasion was present in 42.4% (324/764). Complete resection was achieved in 80.0% of adenomas and subtotal resection, in 20.1%. By multivariate analysis, invasion according to intraoperative findings was associated with the sparsely granulated corticotroph (SGCA, P = .0026) and sparsely granulated somatotroph (SGSA, P = .0103) adenoma type as well as age (P = .0287). Radiographic invasion according to Knosp grades 3A-4 correlated with age (P = .0098), SGCAs (P = .0005), SGSAs (P = .0351), and gonadotroph adenomas (P = .0478). Both criteria of invasion correlated with subtotal resection (P = .0001, respectively). Both intraoperative and radiographic signs of invasive growth are high-risk lesions for incomplete extent of resection and occur more frequently in older patients. A particularly high prevalence of invasion can be found in the SGCA and SGSA types. Cavernous sinus invasion is also more common in gonadotroph adenomas. Usage of the Knosp classification is a valuable preoperative estimation tool.

Enhanced reliability and time efficiency of deep learning-based posterior tibial slope measurement over manual techniques.

Multifaceted factors contribute to inferior outcomes following anterior cruciate ligament (ACL) reconstruction surgery. A particular focus is placed on the posterior tibial slope (PTS). This study introduces the integration of machine learning and artificial intelligence (AI) for efficient measurements of tibial slopes on magnetic resonance imaging images as a promising solution. This advancement aims to enhance risk stratification, diagnostic insights, intervention prognosisand surgical planning for ACL injuries. Images and demographic information from 120 patients who underwent ACL reconstruction surgery were used for this study. An AI-driven model was developed to measure the posterior lateral tibial slope using the YOLOv8 algorithm. The accuracy of the lateral tibial slope, medial tibial slopeand tibial longitudinal axis measurements was assessed, and the results reached high levels of reliability. This study employed machine learning and AI techniques to provide objective, consistentand efficient measurements of tibial slopes on MR images. Three distinct models were developed to derive AI-based measurements. The study results revealed a substantial correlation between the measurements obtained from the AI models and those obtained by the orthopaedic surgeon across three parameters: lateral tibial slope, medial tibial slopeand tibial longitudinal axis. Specifically, the Pearson correlation coefficients were 0.673, 0.850and 0.839, respectively. The Spearman rank correlation coefficients were 0.736, 0.861and 0.738, respectively. Additionally, the interclass correlation coefficients were 0.63, 0.84and 0.84, respectively. This study establishes that the deep learning-based method for measuring posterior tibial slopes strongly correlates with the evaluations of expert orthopaedic surgeons. The time efficiency and consistency of this technique suggest its utility in clinical practice, promising to enhance workflow, risk assessmentand the customization of patient treatment plans. Level III, cross-sectional diagnostic study.

DNA-Mediated Magnetic-Dimer Assembly for Fault-Free Ultra-High-Field Magnetic Resonance Imaging of Tumors.

Ultra-high-field (UHF) magnetic resonance imaging (MRI) stands as a pivotal cornerstone in biomedical imaging, yet the challenge of false imaging persists, constraining its full potential. Despite the development of dual-mode contrast agents improving conventional MRI, their effectiveness in UHF remains suboptimal due to the high magnetic moment, resulting in diminished T1 relaxivity and excessively enhanced T2 relaxivity. Herein, we report a DNA-mediated magnetic-dimer assembly (DMA) of iron oxide nanoparticles that harnesses UHF-tailored nanomagnetism for fault-free UHF-MRI. DMA exhibits a dually enhanced longitudinal relaxivity of 4.42 mM-1·s-1 and transverse relaxivity of 26.23 mM-1·s-1 at 9 T, demonstrating a typical T1-T2 dual-mode UHF-MRI contrast agent. Importantly, DMA leverages T1-T2 dual-modality image fusion to achieve artifact-free breast cancer visualization, effectively filtering interference from hundred-micrometer-level false-positive signals with unprecedented precision. The UHF-tailored T1-T2 dual-mode DMA contrast agents hold promise for elevating the accuracy of MR imaging in disease diagnosis.

Chitosan-modified manganese oxide-conjugated methotrexate nanoparticles delivering 5-aminolevulinic acid as a dual-modal T1-T2* MRI contrast agent in U87MG cell detection.

Glioblastoma multiforme is a highly aggressive form of brain cancer, and early diagnosis plays a pivotal role in improving patient survival rates. In this regard, molecular magnetic resonance imaging has emerged as a promising imaging modality due to its exceptional sensitivity to minute tissue changes and the ability to penetrate deep into the brain. This study aimed to assess the efficacy of a novel contrast agent in detecting gliomas during MRI scans. The contrast agent utilized modified chitosan coating on manganese oxide nanoparticles. The modification included adding methotrexate and 5-aminolevulinic acid (MnO2/CS@5-ALA-MTX) to target cells with overexpressed folate receptors and breaking down excess hydrogen peroxide in tumor tissue, resulting in enhanced signal intensity in T1-weighted MR images but diminished signal intensity in T2*-weighted MR images. The nanosystem was characterized and evaluated in MR imaging, safety, and ability to target cells both in vivo and in vitro. MTX-free nanoparticles (MnO2/CS@5-ALA NPs) had no obvious cytotoxicity on cell lines U87MG and NIH3T3 after 24/48 h at a concentration of up to 160 µgr/mL (cell viability more than 80%). In this system, methotrexate enables tumor targeting and the MnO2/5-ALA improves T1-T2*-weighted MRI. In addition, MRI scans of mice with M109 carcinoma indicated significant tumor uptake and NP capacity to improve the positive contrast effect. This developed MnO2/CS@5-ALA-MTX nanoparticle system may exhibit great potential in the accurate diagnosis of folate receptor over-expressing cancers such as glioblastoma.

Automated glioblastoma patient classification using hypoxia levels measured through magnetic resonance images

IntroductionThe challenge of treating Glioblastoma (GBM) tumors is due to various mechanisms that make the tumor resistant to radiation therapy. One of these mechanisms is hypoxia, and therefore, determining the level of hypoxia can improve treatment planning and initial evaluation of its effectiveness in GBM. This study aimed to design an intelligent system to classify glioblastoma patients based on hypoxia levels obtained from magnetic resonance images with the help of an artificial neural network (ANN).Material and methodMR images and PET measurements were available for this study. MR images were downloaded from the Cancer Imaging Archive (TCIA) database to classify glioblastoma patients based on hypoxia. The images in this database were prepared from 27 patients with glioblastoma on T1W + Gd, T2W-FLAIR, and T2W. Our designed algorithm includes various parts of pre-processing, tumor segmentation, feature extraction from images, and matching these features with quantitative parameters related to hypoxia in PET images. The system’s performance is evaluated by categorizing glioblastoma patients based on hypoxia.ResultsThe results of classification with the artificial neural network (ANN) algorithm were as follows: the highest sensitivity, specificity, and accuracy were obtained at 86.71, 85.99 and 83.17%, respectively. The best specificity was related to the T2W-EDEMA image with the tumor to blood ratio (TBR) as a hypoxia parameter. T1W-NECROSIS image with the TBR parameter also showed the highest sensitivity and accuracy.ConclusionThe results of the present study can be used in clinical procedures before treating glioblastoma patients. Among these treatment approaches, we can mention the radiotherapy treatment design and the prescription of effective drugs for the treatment of hypoxic tumors.

Variations in knee cartilage thickness: Fully automatic three-dimensional analysis of MRIs from five manufacturers

BackgroundMeasurements of knee cartilage thickness derived from MR images are attractive biomarkers for osteoarthritis research. Although some cross-sectional multivendor studies exist, none have employed fully automatic three-dimensional MRI analysis. Our objective was to evaluate the variations in knee cartilage thickness measurements obtained using automated methods and MRI instruments from five different vendors. MethodsThe subjects were 10 healthy volunteers aged 22–60 years. MRI models with 3 Tesla strength from five different companies were used. Cartilage thickness was quantified fully automatically for seven regions. We hypothesized that “the MRI model influences cartilage thickness measurements.” Inter-measurement error, defined as the absolute difference between the targeted and median thicknesses determined by the five MRI models, was analyzed using histograms. The factors generating the largest inter-measurement error were also examined. ResultsNo exceptional trends attributable to a specific instrument model were observed, and the p-value from the Kruskal–Wallis test exceeded 0.05 in all seven regions. Therefore, the study hypothesis was rejected. Of the 350 measurements, the inter-measurement error was ≤0.05 mm in 53 %, ≤0.10 mm in 75 %, and ≤0.20 mm in 95 %. Analysis of the medial tibial cartilage, which had the largest inter-measurement error, revealed mis-extraction of synovial fluid as cartilage. ConclusionsThe choice of MRI model did not influence cartilage thickness measurements. Overall, 95 % of the inter-measurement errors were within 0.20 mm. The greatest error resulted from mis-extracting synovial fluid as cartilage.

A predictive model using MRI and clinicopathologic features for breast cancer recurrence in young women treated with upfront surgery.

To identify preoperative breast MR imaging and clinicopathological variables related to recurrence and develop a risk prediction model for recurrence in young women with breast cancer treated with upfront surgery. This retrospective study analyzed 438 consecutive women with breast cancer aged 35 years or younger between January 2007 and December 2016. Breast MR images before surgery were independently reviewed by breast radiologists blinded to patient outcomes. The clinicopathological data including patient demographics, clinical features, and tumor characteristics were reviewed. Univariate and multivariate logistic regression analyses were used to identify the independent factors associated with recurrence. The risk prediction model for recurrence was developed, and the discrimination and calibration abilities were assessed. Of 438 patients, 95 (21.7%) developed recurrence after a median follow-up of 65 months. Tumor size at MR imaging (HR = 1.158, p = 0.006), multifocal or multicentric disease (HR = 1.676, p = 0.017), and peritumoral edema on T2WI (HR = 2.166, p = 0.001) were identified as independent predictors of recurrence, while adjuvant endocrine therapy (HR = 0.624, p = 0.035) was inversely associated with recurrence. The prediction model showed good discrimination ability in predicting 5-year recurrence (C index, 0.707 in the development cohort; 0.686 in the validation cohort) and overall recurrence (C index, 0.699 in the development cohort; 0.678 in the validation cohort). The calibration plot demonstrated an excellent correlation (concordance correlation coefficient, 0.903). A prediction model based on breast MR imaging and clinicopathological features showed good discrimination to predict recurrence in young women with breast cancer treated with upfront surgery, which could contribute to individualized risk stratification. Our prediction model, incorporating preoperative breast MR imaging and clinicopathological features, predicts recurrence in young women with breast cancer undergoing upfront surgery, facilitating personalized risk stratification and informing tailored management strategies. Younger women with breast cancer have worse outcomes than those diagnosed at more typical ages. The described prediction model showed good discrimination performance in predicting 5-year and overall recurrence. Incorporating better risk stratification tools in this population may help improve outcomes.

NBEST: Deep-learning-based non-human primates Brain Extraction and Segmentation Toolbox across ages, sites and species

Accurate processing and analysis of non-human primate (NHP) brain magnetic resonance imaging (MRI) serves an indispensable role in understanding brain evolution, development, aging, and diseases. Despite the accumulation of diverse NHP brain MRI datasets at various developmental stages and from various imaging sites/scanners, existing computational tools designed for human MRI typically perform poor on NHP data, due to huge differences in brain sizes, morphologies, and imaging appearances across species, sites, and ages, highlighting the imperative for NHP-specialized MRI processing tools. To address this issue, in this paper, we present a robust, generic, and fully automated computational pipeline, called non-human primates Brain Extraction and Segmentation Toolbox (nBEST), whose main functionality includes brain extraction, non-cerebrum removal, and tissue segmentation. Building on cutting-edge deep learning techniques by employing lifelong learning to flexibly integrate data from diverse NHP populations and innovatively constructing 3D U-NeXt architecture, nBEST can well handle structural NHP brain MR images from multi-species, multi-site, and multi-developmental-stage (from neonates to the elderly). We extensively validated nBEST based on, to our knowledge, the largest assemblage dataset in NHP brain studies, encompassing 1,469 scans with 11 species (e.g., rhesus macaques, cynomolgus macaques, chimpanzees, marmosets, squirrel monkeys, etc.) from 23 independent datasets. Compared to alternative tools, nBEST outperforms in precision, applicability, robustness, comprehensiveness, and generalizability, greatly benefiting downstream longitudinal, cross-sectional, and cross-species quantitative analyses. We have made nBEST an open-source toolbox (https://github.com/TaoZhong11/nBEST) and we are committed to its continual refinement through lifelong learning with incoming data to greatly contribute to the research field.

P.109 Machine learning based approach to improving the prediction of neurological deterioration in mild Degenerative Cervical Myelopathy

Background: Degenerative cervical myelopathy (DCM) is the most common form of atraumatic spinal cord injury globally, yet clinical guidelines remain unclear on surgical recommendations for patients with mild forms of DCM. This is in part due to limitations in current MR imaging interpretation and complex mechanisms of neurological deterioration. Supervised machine learning (ML) models can help to identify clinical and imaging indicators of deterioration within mild DCM patients. Methods: 127 MRI scans (T2w, Diffusion Tensor Imaging, and Magnetization transfer scans) accompanied by a series of clinical tests underwent a semi-automated analysis to derive quantitative metrics. Random forest classifier, Support Vector Machine, and Logistic Regression models were trained and tested to predict 6-month neurological deterioration within patients. Results: The ML models performed, on average, better than previous studies with a balanced accuracy ranging between 70-75%. “Advanced” imaging metrics such as diffusion tensor imaging and magnetization transfer scans played an important role in improving model accuracy but only when used near the maximally compressed disc level, suggesting that limited yet targetted imaging metrics support ML model performance. Conclusions: The inclusion of specific, targeted imaging and clinical metrics support ML model performance in predicting neurological deterioration within mild DCM patients.

A Comprehensive Evaluation of Deep Learning Models on Knee MRIs for the Diagnosis and Classification of Meniscal Tears: A Systematic Review and Meta-Analysis.

This study delves into the cutting-edge field of deep learning techniques, particularly deep convolutional neural networks (DCNNs), which have demonstrated unprecedented potential in assisting radiologists and orthopedic surgeons in precisely identifying meniscal tears. This research aims to evaluate the effectiveness of deep learning models in recognizing, localizing, describing, and categorizing meniscal tears in magnetic resonance images (MRIs). This systematic review was rigorously conducted, strictly following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. Extensive searches were conducted on MEDLINE (PubMed), Web of Science, Cochrane Library, and Google Scholar. All identified articles underwent a comprehensive risk of bias analysis. Predictive performance values were either extracted or calculated for quantitative analysis, including sensitivity and specificity. The meta-analysis was performed for all prediction models that identified the presence and location of meniscus tears. This study's findings underscore that a range of deep learning models exhibit robust performance in detecting and classifying meniscal tears, in one case surpassing the expertise of musculoskeletal radiologists. Most studies in this review concentrated on identifying tears in the medial or lateral meniscus and even precisely locating tears-whether in the anterior or posterior horn-with exceptional accuracy, as demonstrated by AUC values ranging from 0.83 to 0.94. Based on these findings, deep learning models have showcased significant potential in analyzing knee MR images by learning intricate details within images. They offer precise outcomes across diverse tasks, including segmenting specific anatomical structures and identifying pathological regions. Contributions: This study focused exclusively on DL models for identifying and localizing meniscus tears. It presents a meta-analysis that includes eight studies for detecting the presence of a torn meniscus and a meta-analysis of three studies with low heterogeneity that localize and classify the menisci. Another novelty is the analysis of arthroscopic surgery as ground truth. The quality of the studies was assessed against the CLAIM checklist, and the risk of bias was determined using the QUADAS-2 tool.

Magnetic resonance image reconstruction based on image decomposition constrained by total variation and tight frame.

Magnetic resonance imaging (MRI) is a commonly used tool in clinical medicine, but it suffers from the disadvantage of slow imaging speed. To address this, we propose a novel MRI reconstruction algorithm based on image decomposition to realize accurate image reconstruction with undersampled k-space data. In our algorithm, the MR images to be recovered are split into cartoon and texture components utilizing image decomposition theory. Different sparse transform constraints are applied to each component based on their morphological structure characteristics. The total variation transform constraint is used for the smooth cartoon component, while the L0 norm constraint of tight frame redundant transform is used for the oscillatory texture component. Finally, an alternating iterative minimization approach is adopted to complete the reconstruction. Numerous numerical experiments are conducted on several MR images and the results consistently show that, compared with the existing classical compressed sensing algorithm, our algorithm significantly improves the peak signal-to-noise ratio of the reconstructed images and preserves more image details. Our algorithm harnesses the sparse characteristics of different image components to reconstruct MR images accurately with highly undersampled data. It can greatly accelerate MRI speed and be extended to other imaging reconstruction fields.

Differentiating Low-Grade from High-Grade Intracranial Ependymomas: Comparison of Dynamic Contrast-Enhanced MRI and Diffusion-Weighted Imaging.

The aim of this study was to determine the diagnostic value of fractional plasma volume derived from dynamic contrast-enhanced perfusion MR imaging versus ADC, obtained from DWI in differentiating between grade 2 (low-grade) and grade 3 (high-grade) intracranial ependymomas. A hospital database was created for the period from January 2013 through June 2022, including patients with histologically-proved ependymoma diagnosis with available dynamic contrast-enhanced MR imaging. Both dynamic contrast-enhanced perfusion and DWI were performed on each patient using 1.5T and 3T scanners. Fractional plasma volume maps and ADC maps were calculated. ROIs were defined by a senior neuroradiologist manually by including the enhancing tumor on every section and conforming a VOI to obtain the maximum value of fractional plasma volume (Vpmax) and the minimum value of ADC (ADCmin). A Mann-Whitney U test at a significance level of corrected P = .01 was used to evaluate the differences. Additionally, receiver operating characteristic curve analysis was applied to assess the sensitivity and specificity of Vpmax and ADCmin values. A total of 20 patients with ependymomas (10 grade 2 tumors and 10 grade 3 tumors) were included. Vpmax values for grade 3 ependymomas were significantly higher (P < .002) than those for grade 2. ADCmin values were overall lower in high-grade lesions. However, no statistically significant differences were found (P = .12114). As a dynamic contrast-enhanced perfusion MR imaging metric, fractional plasma volume can be used as an indicator to differentiate grade 2 and grade 3 ependymomas. Dynamic contrast-enhanced perfusion MR imaging plays an important role with high diagnostic value in differentiating low- and high-grade ependymoma.