Magnetic Resonance Imaging Sequences Research Articles

Differentiating the Presence of Brain Stroke Types in MR Images using CNN Architecture.

Stroke is reportedly the biggest cause of death and disability in the world, according to the World Health Organisation (WHO). The severity of a stroke can be lowered by recognising the many stroke warning indicators early on. Using CNN, the primary goal of this study is to predict the likelihood that a brain stroke would develop at an early stage. The novelty of the proposed work is to acquire models that can accurately differentiate between stroke and no-stroke (normal) cases using MR Imaging sequences like DWI, SWI, GRE and T2 FLAIR aiding in timely diagnosis and treatment decisions. A dataset comprising real time MRI scans of patients with stroke and no-stroke conditions was collected and preprocessed for model training. The preprocessing involves standardizing the resolution of the images, normalizing pixel values, and augmenting the dataset to enhance model generalization. The ResNet, DenseNet, EfficientNet, and VGG16 architectures were implemented and trained on the preprocessed dataset. The training process involved optimizing model parameters using stochastic gradient descent and minimizing the loss function. The results demonstrate promising performance across all models by obtaining an accuracy of 98% for ResNet, DenseNet and EfficientNet, while 97% for VGG16 in differentiating the stroke using real time MRI data. The proposed work explored the effectiveness of CNN models, including ResNet, DenseNet, EfficientNet, and VGG16, for the differentiation of stroke and no-stroke cases. The models were trained and evaluated using a real-time dataset of brain MR Images. The obtained accuracies highlight the potential of CNN models in accurately differentiating between stroke and non-stroke cases.

A New Method Used to Enhance the SPAIR Image of the Spine MRI.

Magnetic resonance imaging (MRI) is a handy diagnostic tool for orthopedic disorders, particularly spinal and joint diseases. The lumbar intervertebral disc is visible in the T1 and T2 weight sequences of the spine MRI, which aids in diagnosing lumbar disc herniation, lumbar spine tuberculosis, lumbar spine tumors, and other conditions. The lumbar intervertebral disc cannot be seen accurately in the Spectral Attenuated Inversion Recovery (SPAIR) due to weaknesses in the fat and frequency offset parameters, which is not conducive to developing the intelligence diagnosis model of medical image. In order to solve this problem, we propose a composite framework, which is first to use the contrast limited adaptive histogram equalization (CLAHE) method to enhance the SPAIR image contrast of the spine MRI and then use the non-local means method to remove the noise of the image to ensure that the image contrast is uniform without losing details. We employ the Information Entropy (IE), Peak signal-to-noise ratio (PSNR), and feature similarity index measure (FSIM) to quantify image quality after enhancement by the composite framework. The outcomes of the experiments' output images and quantitative data indicate that our composite framework is better than others.

Infratentorial superficial siderosis: report of six cases and review of the literature.

To investigate the etiology, clinical manifestations, imaging features, and treatment of patients with infratentorial superficial siderosis (iSS), enhance clinicians' comprehension of this rare disease, and conduct oral deferiprone intervention and subsequent monitoring. Six patients diagnosed with iSS based on magnetic resonance imaging (MRI) and susceptibility weighted imaging (SWI) were enrolled from 2021 to 2023 at the First Affiliated Hospital of Fujian Medical University. Their clinical datas were summarized, and the etiology and imaging characteristics were analyzed. Follow-up was conducted through telephone or outpatient visits. Among the 6 patients, there were 3 males and 3 females. The onset age ranged from 35 to 71 years, with an average onset age of 53 years. The clinical symptoms mainly included acoustic disturbances (6/6), gait imbalance (6/6), dysolfactory (6/6), cognitive impairment (2/6), epilepsy (2/6), and pyramidal tract sign (2/6). Evidence of superficial siderosis was observed on MRI across the cortex, brainstem, cerebellum, and spinal cord in all patients. T2-space sequence MRI revealed two instances of dural tear. During the follow-up period ranging from 1 month to 3 years, three patients who received oral deferiprone treatment showed improvement, whereas the remaining three patients who declined deferiprone treatment demonstrated progression. The primary clinical manifestations of iSS include bilateral sensorineural hearing disturbances, progressive cerebellar ataxia, and spinal cord lesions. The key diagnostic criteria involve the presence of linear hypointensity on T2-WI in the surface region of the nervous system. Dural tear caused by various factors is considered to be the most common cause of iSS, and its treatment mainly involves surgical intervention for hemorrhagic primary diseases as well as pharmacotherapy with deferiprone.

The magnetization transfer ratio of the post-mortem canine intervertebral disk is positively correlated to Pfirrmann grading on high field 3.0T MRI: a pilot study.

Intervertebral disk (IVD) degeneration usually occurs earlier in chondrodystrophic dog breeds than in other breeds. Spinal cord compression secondary to IVD degeneration is the most common cause of myelopathy in these dogs. Standard magnetic resonance imaging (MRI) sequences permit the identification of IVD degeneration and its consequences on adjacent neurological structures. In human medicine, quantitative MRI sequences, such as magnetization transfer ratio (MTR) sequences, are developed and used to detect early IVD degeneration. This prospective randomized post-mortem comparative study aimed to evaluate the correlation between a qualitative Pfirrmann MRI grading and the MTR values of the IVD in chondrodystrophic dogs. Vertebral columns of eight canine cadavers were frozen and thawed prior to imaging with T2-weighted and MTR sequences using a 3.0 T high-field MRI. These sequences were reviewed by two observers. A Spearman correlation coefficient was calculated in order to compare the MTR values with the Pfirrmann grade. Pearson correlation coefficients were calculated to evaluate the inter-observer agreement of the delineation of the region of interest (ROI) around the NP and the MTR values. A Wilcoxon-Mann-Whitney test was used to conclude on the significance of the correlation between the MTR values and the Pfirrmann grades. There were 138 intervertebral disks analyzed: 29/138 (21.0%) IVD were grade I, 74/138 (53.6%) grade II, and 35/138 (25.4%) grade III. No grades IV and V were present in this study. Inter-observer agreement for delineation of IVD ROI was fair (r = 0.54) but inter-observer agreement of mean MTR value within the ROI was very good (r = 0.89). Mean MTR values were 16.459% (10.0305-21.0950%) for grade I, 18.888% (10.0750-27.2400%) for grade II, and 22.813% (12.5700-31.7600%) for grade III. The mean MTR value was significantly different between each Pfirrmann grade: between grades I and II (p < 0.005), grades II and III (p < 0.05), and grades I and III (p < 0.005). There was a significant moderate positive correlation between Pfirrmann grading and mean MTR values (r = 0.516). The magnetization transfer ratio seems to be an objective method to detect early intervertebral disk degeneration via quantitative analysis.

Static and dynamic responses to hyperoxia of normal placenta across gestation with T2*-weighted MRI sequences.

T2*-weighted magnetic resonance imaging (MRI) sequences have been identified as non-invasive tools with which to study placental oxygenation in vivo. This study aimed to use these to investigate both static and dynamic responses to hyperoxia of the normal placenta across gestation. We conducted a single-center prospective study including 52 uncomplicated pregnancies. Two T2*-weighted sequences (T2* relaxometry) were performed, one before and one after maternal hyperoxia. The distribution of placental T2* values was modeled by fitting a gamma probability density function (T2* ), describing the structure of the histogram using the mean T2* value, the shape parameter (α) and the rate (β). A dynamic acquisition (blood-oxygen-level-dependent (BOLD) MRI) was also performed before and during maternal oxygen supply, until placental oxygen saturation had been achieved. The signal change over time was modeled using a sigmoid function, to determine the intensity of enhancement (ΔBOLD (% with respect to baseline)), a temporal variation coefficient (λ (min-1), controlling the slope of the curve) and the maximum steepness (Vmax (% of placental enhancement/min)). The histogram analysis of the T2* values in normoxia showed a whole-placenta variation, with a decreasing linear trend in the mean T2* value (Pearson's correlation coefficient (R) = -0.83 (95% CI, -0.9 to -0.71), P < 0.001), along with an increasingly peaked and narrower distribution of T2* values with advancing gestation. After maternal hyperoxia, the mean T2* ratios (mean T2*hyperoxia/mean T2*baseline) were positively correlated with gestational age, while the other histogram parameters remained stable, suggesting a translation of the histogram towards higher values with a similar appearance after maternal hyperoxia. ΔBOLD showed a non-linear increase across gestation. Conversely, λ showed an inverted trend across gestation, with a weaker correlation (R = -0.33 (95% CI, -0.58 to -0.02), P = 0.04, R2 = 0.1). As a combination of ΔBOLD and λ, the changes in Vmax throughout gestation were influenced mainly by the changes in ΔBOLD and showed a positive non-linear correlation with gestational age. Our results suggest that the decrease in the T2* placental signal as gestation progresses does not reflect placental dysfunction. The BOLD dynamic signal change is representative of a free-diffusion model of oxygenation and highlights the increasing differences in oxygen saturation between mother and fetus as gestation progresses (ΔBOLD) and in the placental permeability to oxygen (λ). © 2024 International Society of Ultrasound in Obstetrics and Gynecology.

Association of Rare NOTCH3 Variants With Prevalent and Incident Stroke and Dementia in the General Population.

It is uncertain whether rare NOTCH3 variants are associated with stroke and dementia in the general population and whether they lead to alterations in cognitive function. This study aims to determine the associations of rare NOTCH3 variants with prevalent and incident stroke and dementia, as well as cognitive function changes. In the prospective community-based Shunyi Study, a total of 1007 participants were included in the baseline analysis. For the follow-up analysis, 1007 participants were included in the stroke analysis, and 870 participants in the dementia analysis. All participants underwent baseline brain magnetic resonance imaging, carotid ultrasound, and whole exome sequencing. Rare NOTCH3 variants were defined as variants with minor allele frequency <1%. A total of 137 rare NOTCH3 carriers were enrolled in the baseline study. At baseline, rare NOTCH3 variant carriers had higher rates of stroke (8.8% versus 5.6%) and dementia (2.9% versus 0.8%) compared with noncarriers. After adjustment for associated risk factors, the epidermal growth factor-like repeats (EGFr)-involving rare NOTCH3 variants were associated with a higher risk of prevalent stroke (odds ratio [OR], 2.697 [95% CI, 1.266-5.745]; P=0.040) and dementia (OR, 8.498 [95% CI, 1.727-41.812]; P=0.032). After 5 years of follow-up, we did not find that the rare NOTCH3 variants increased the risk of incident stroke and dementia. There was no statistical difference in the change in longitudinal cognitive scale scores. Rare NOTCH3 EGFr-involving variants are genetic risk factors for stroke and dementia in the general Chinese population.

Effects of high-intensity training on fatty infiltration in paraspinal muscles in elderly males with osteosarcopenia – the randomized controlled FrOST study

BackgroundOsteosarcopenia is a common geriatric syndrome with an increasing prevalence with age, leading to secondary diseases and complex consequences such as falls and fractures, as well as higher mortality and frailty rates. There is a great need for prevention and treatment strategies.MethodsIn this analysis, we used magnetic resonance imaging (MRI) data from the randomised controlled FrOST trial, which enrolled community-dwelling osteosarcopenic men aged > 72 years randomly allocated to 16 months of twice-weekly high-intensity resistance training (HIRT) or a non-training control group. MR Dixon imaging was used to quantify the effects of HIRT on muscle fat infiltration in the paraspinal muscles, determined as changes in muscle tissue, fat faction and intermuscular adipose tissue (IMAT) in the erector spinae and psoas major muscles. Intention-to-treat analysis with multiple imputation was used to analyse the data set.ResultsAfter 16 months of intervention, 15 men from the HIRT and 16 men from the CG were included in the MRI analysis. In summary, no positive effects on the fat infiltration of the erector spinae and psoas major muscles were observed.ConclusionsThe previously reported positive effects on lumbar spine bone mineral density (BMD) suggest that mechanotransduction induces tropic effects on bone, but that fat infiltration of the erector spinae and psoas major muscles are either irreversible or, for some unknown reason, resistant to exercise. Because of the beneficial effects on spinal BMD, HIRT is still recommended in osteosarcopenic older men, but further research is needed to confirm appropriate age-specific training exercises for the paraspinal muscles. The potential of different MRI sequences to quantify degenerative and metabolic changes in various muscle groups must be better characterized.Trial registrationsFrOST was approved by the University Ethics Committee of the Friedrich-Alexander University of Erlangen-Nürnberg (number 67_15b and 4464b) and the Federal Office for Radiation Projection (BfS, number Z 5–2,246,212 – 2017–002). Furthermore, it fully complies with the Declaration of Helsinki and is registered at ClinicalTrials.gov: NCT03453463 (05/03/2018). JAMA 310:2191–2194, 2013.

Robot-Assisted Lumbar Pedicle Screw Placement Based on 3D Magnetic Resonance Imaging.

Human Cadaveric Study. This study aims to explore the feasibility of using preoperative magnetic resonance imaging (MRI), zero-time-echo (ZTE) and spoiled gradient echo (SPGR), as source data for robotic-assisted spine surgery and assess the accuracy of pedicle screws. Zero-time-echo and SPGR MRI scans were conducted on a human cadaver. These images were manually post-processed, producing a computed tomography (CT)-like contrast. The Mazor X robot was used for lumbar pedicle screw-place navigating of MRI. The cadaver underwent a postoperative CT scan to determine the actual position of the navigated screws. Ten lumbar pedicle screws were robotically navigated of MRI (4 ZTE; 6 SPGR). All MR-navigated screws were graded A on the Gertzbein-Robbins scale. Comparing preoperative robotic planning to postoperative CT scan trajectories: The screws showed a median deviation of overall 0.25mm (0.0; 1.3), in the axial plane 0.27mm (0.0; 1.3), and in the sagittal plane 0.24mm (0.0; 0.7). This study demonstrates the first successful registration of MRI sequences, ZTE and SPGR, in robotic spine surgery here used for intraoperative navigation of lumbar pedicle screws achieving sufficient accuracy, showcasing potential progress toward radiation-free spine surgery.

Deep learning algorithm-based multimodal MRI radiomics and pathomics data improve prediction of bone metastases in primary prostate cancer

PurposeBone metastasis is a significant contributor to morbidity and mortality in advanced prostate cancer, and early diagnosis is challenging due to its insidious onset. The use of machine learning to obtain prognostic information from pathological images has been highlighted. However, there is a limited understanding of the potential of early prediction of bone metastasis through the feature combination method from various sources. This study presents a method of integrating multimodal data to enhance the feasibility of early diagnosis of bone metastasis in prostate cancer.Methods and materialsOverall, 211 patients diagnosed with prostate cancer (PCa) at Gansu Provincial Hospital between January 2017 and February 2023 were included in this study. The patients were randomized (8:2) into a training group (n = 169) and a validation group (n = 42). The region of interest (ROI) were segmented from the three magnetic resonance imaging (MRI) sequences (T2WI, DWI, and ADC), and pathological features were extracted from tissue sections (hematoxylin and eosin [H&E] staining, 10 × 20). A deep learning (DL) model using ResNet 50 was employed to extract deep transfer learning (DTL) features. The least absolute shrinkage and selection operator (LASSO) regression method was utilized for feature selection, feature construction, and reducing feature dimensions. Different machine learning classifiers were used to build predictive models. The performance of the models was evaluated using receiver operating characteristic curves. The net clinical benefit was assessed using decision curve analysis (DCA). The goodness of fit was evaluated using calibration curves. A joint model nomogram was eventually developed by combining clinically independent risk factors.ResultsThe best prediction models based on DTL and pathomics features showed area under the curve (AUC) values of 0.89 (95% confidence interval [CI], 0.799–0.989) and 0.85 (95% CI, 0.714–0.989), respectively. The AUC for the best prediction model based on radiomics features and combining radiomics features, DTL features, and pathomics features were 0.86 (95% CI, 0.735–0.979) and 0.93 (95% CI, 0.854–1.000), respectively. Based on DCA and calibration curves, the model demonstrated good net clinical benefit and fit.ConclusionMultimodal radiomics and pathomics serve as valuable predictors of the risk of bone metastases in patients with primary PCa.

Comprehensive evaluation of magnetic resonance imaging sequences for signal intensity based assessment of anterior cruciate ligament healing following surgical treatment.

Normalized signal intensity (SI) obtained from magnetic resonance imaging (MRI) has been used to track anterior cruciate ligament (ACL) postoperative remodeling. We aimed to assess the effect of MRI sequence (PD: proton density-weighted; T2: T2-weighted; CISS: constructive interference in steady state) on postoperative changes in healing ACLs/grafts. We hypothesized that CISS is better at detecting longitudinal SI and texture changes of the healing ACL/graft compared to the common clinical sequences (PD and T2). MR images of patients who underwent ACL surgery were evaluated and separated into groups based on surgical procedure (Bridge-Enhanced ACL Repair (BEAR; n = 50) versus ACL reconstruction (ACLR; n = 24)). CISS images showed decreasing SI across all timepoints in both the BEAR and ACLR groups (p < 0.01), PD and T2 images showed decreasing SI in the 6-to-12- and 12-to-24-month postoperative timeframes in the BEAR group (p < 0.02), and PD images additionally showed decreasing SI between 6- and 24-months postoperation in the ACLR group (p = 0.02). CISS images showed texture changes in both the BEAR and ACLR groups, showing increases in energy and decreases in entropy in the 6-to-12- and 6-to-24-month postoperative timeframes in the BEAR group (p 0.04), and increases in energy, decreases in entropy, and increases in homogeneity between 6 and 24 months postoperation in the ACLR group (p < 0.04). PD images showed increases in energy and decreases in entropy between 6- and 24-months postoperation in the ACLR group (p < 0.008). Finally, CISS was estimated to require a smaller sample size than PD and T2 to detect SI differences related to postoperative remodeling.

Comparing the Efficacy of a Combination of Diffusion-Weighted Imaging and T2-STIR (Short Tau Inversion Recovery) Imaging With Contrast-Enhanced MRI in the Evaluation of Perianal Fistula.

Perianal fistula is clinically diagnosed and commonly characterized using magnetic resonance imaging (MRI). Diffusion-weighted imaging (DWI) and T2-weighted imaging are emerging techniques that can obviate the need for contrast injection in cases where contrast administration is not feasible or contraindicated. The main objective of our study was to compare the efficacy of the combination of DWI and T2 STIR (short tau inversion recovery) imaging with contrast-enhanced MRI for the diagnosis and characterization of perianal fistula. Sixty-nine patients with clinical perianal fistula with at least one external opening were evaluated with DWI, T2 STIR, and contrast MRI. A comparative cross-sectional study was conducted in the Department of Radiodiagnosis and Imaging, All India Institute of Medical Sciences, Bhopal, India. The chi-square test was done to find the association between categorical variables. The Kappa test was done to estimate the agreement between two different tests in measuring the outcome. The validity of tests was measured using sensitivity, specificity, positive predictive value, negative predictive value, and accuracy. The combination of DWI and T2 STIR is equivalent to contrast-enhanced MRI in the evaluation of primary and complicated perianal fistula. The combination of DWI and T2 STIR is superior to DWI alone in the classification and characterization of perianal fistula. However, DWI is superior to T2 STIR in differentiating perianal inflammation with abscess from perianal inflammation without abscess and can be used as an alternative to post-contrast fat-suppressed T1-WI in the detection of perianal abscesses and disease activity. DWI can be used as an adjunct to T2 STIR, and the combination of DWI and T2 STIR can replace the post-contrast fat-suppressed T1 MRI sequence in the classification and characterization of perianal fistula.

Brain tumor grade classification using the ConvNext architecture.

Brain tumor grade is an important aspect of brain tumor diagnosis and helps to plan for treatment. Traditional methods of diagnosis, including biopsy and manual examination of medical images, are either invasive or may result in inaccurate diagnoses. This study proposes a brain tumor grade classification technique using a modern convolutional neural network (CNN) architecture called ConvNext that inputs magnetic resonance imaging (MRI) data. Deep learning-based techniques are replacing invasive procedures for consistent, accurate, and non-invasive diagnosis of brain tumors. A well-known challenge of using deep learning architectures in medical imaging is data scarcity. Modern-day architectures have huge trainable parameters and require massive datasets to achieve the desired accuracy and avoid overfitting. Therefore, transfer learning is popular among researchers using medical imaging data. Recently, transformer-based architectures have surpassed CNNs for image data. However, recently proposed CNNs have achieved superior accuracy by introducing some tweaks inspired by vision transformers. This study proposed a technique to extract features from the ConvNext architecture and feed these features to a fully connected neural network for final classification. The proposed study achieved state-of-the-art performance on the BraTS 2019 dataset using pre-trained ConvNext. The best accuracy of 99.5% was achieved when three MRI sequences were input as three channels of the pre-trained CNN. The study demonstrated the efficacy of the representations learned by a modern CNN architecture, which has a higher inductive bias for the image data than vision transformers for brain tumor grade classification.

Diagnostic Contribution of Additional Sequences to the Evaluation of Cord Lesions in Patients with Cervical Spinal Multiple Sclerosis in Turkey: A Retrospective Study.

Multiple Sclerosis (MS) is the most common cause of non-traumatic disability in young adults. Spinal cord involvement is observed in 55-75% of patients with MS. To identify the strengths and shortcomings of sagittal phase-sensitive inversion recovery (PSIR), sagittal proton density/T2-weighted (PD/T2W), and axial turbo inversion recovery magnitude (TIRM) sequences in the detection of cervical MS plaques by comparing with routine sequences (axial and sagittal T2W, sagittal T1W, sagittal TIRM, fat-suppressed contrast T1W) and therefore determine their diagnostic contributions. A total of 48 patients in whom additional magnetic resonance imaging (MRI) sequences were obtained for the diagnosis of cervical MS were retrospectively identified and included in the study. A total of 111 MS plaques were analyzed in terms of visibility, number, size, border sharpness, and intensity ratio based on the routine and additional MRI sequences. The evaluation of the images was independently undertaken by two radiologists. The highest visibility was provided by sagittal PSIR, sagittal TIRM, and axial TIRM sequences (P < 0.05 for all additional sequences). Seven lesions in PD/T2W and four lesions in axial T2W sequences were unable to be detected. Lesions seen in sagittal and axial TIRM sequences were larger than the others. The sharpest borders were determined in the axial TIRM sequence, and the most diffuse borders in the PD/T2W sequence. In intensity ratio, the sagittal PSIR sequence revealed the most significant contrast difference. The sagittal PSIR sequence may improve the detection of cervical MS plaques due to the improved visibility and intensity ratios. The axial TIRM sequence may be more useful than routine axial T2W in the evaluation of visibility, border sharpness, and size measurement of MS plaques.

Narrative review of magnetic resonance imaging in quantifying liver iron load.

To summarize the research progress of magnetic resonance imaging (MRI) in quantifying liver iron load. To summarize the current status and progress of MRI technology in the quantitative study of liver iron load through reviewing the relevant literature at home and abroad. Different MRI sequence examination techniques have formed a series of non-invasive methods for the examination of liver iron load. These techniques have important clinical significance in the imaging diagnosis of liver iron load. So far, the main MRI methods used to assess liver iron load are: signal intensity measurement method (signal intensity, SI) [signal intensity ratio (SIR) and difference in in-phase and out-of-phase signal intensity], T2/R2 measurement (such as FerriScan technique), ultra-short echo time (UTE) imaging technique, and susceptibility weighted imaging (including conventional susceptibility weighted imaging) (SWI), quantitative susceptibility mapping (QSM), T2*/R2* measurement, Dixon and its derivative techniques. MRI has become the first choice for the non-invasive examination of liver iron overload, and it is helpful to improve the early detection of liver injury, liver fibrosis, liver cirrhosis and liver cancer caused by liver iron overload.

Guess What Is in My Brain.

Magnetic resonance imaging (MRI) of the brain is an important diagnostic tool used by neurologists. This article explores the workup and management for a patient with a brain lesion and highlights the importance of neuroimaging. Similarities and differences in MRI findings for meningioma, central nervous system lymphoma, and glioblastomas are discussed, along with common MRI sequences and their utility.

Correlation of Static and Dynamic Magnetic Resonance Imaging Sequences in Evaluation of Temporomandibular Joint Dysfunction

Correlation of Static and Dynamic Magnetic Resonance Imaging Sequences in Evaluation of Temporomandibular Joint Dysfunction

Measurement of TT-TG can change with sequential MRIs due to variations in tibiofemoral rotation in patellofemoral instability patients.

There are various anatomic risk factors for patellofemoral instability (PFI) that help guide surgical treatment, including the tibial tubercle to trochlear groove (TT-TG) distance. However, no study has analysed the temporal changes in TT-TG prior to surgical intervention. This study sought to understand the variations in TT-TG over time for pediatric patients suffering from PFI prior to surgical intervention. The authors hypothesised that the TT-TG would substantially change between time points. Patients undergoing medial patellofemoral ligament (MPFL) reconstruction between 2014 and 2019 by one of two fellowship-trained orthopaedic surgeons were identified. Patients were included if they had two preoperative magnetic resonance imaging (MRI) performed on the same knee within 7.5 months of each other prior to any surgical intervention and had an initial TT-TG greater than 10 mm. After considering 251 patients for inclusion, 21 patients met the final inclusion criteria. The mean age was 14.5 ± 2.5 years and 61.9% were female. TT-TG was initially noted to be 15.1 ± 1.8 mm. At mean time aftersequential MRIs of 5.0 ± 1.9 months, TT-TG was noted to be 16.7 ± 3.2 mm. The differences between initial and subsequent TT-TG ranged from a 21.2% decrease to a 61.1% increase, with a mean difference of an 11.3% increase. Comparison between initial and subsequent TT-TG values demonstrated a significant difference (p = 0.017). Change in tibiofemoral rotation ranged from -9.2° to 7.5°. When comparing the change in TT-TG to change in tibiofemoral rotation, a significant correlation was found (p = 0.019). Despite only a mean time between MRIs of 5 months, variations in TT-TG ranged from a decrease of 21.2% to an increase of 61.1%. The significant relationship between the changes in TT-TG and changes in tibiofemoral rotation between MRIs suggest that TT-TG measurements may vary due to variations in tibiofemoral rotation at the time of individual MRIs. Level IV.

Time-Saving 3D MR Imaging Protocols with Millimeter and Submillimeter Isotropic Spatial Resolution for Face and Neck Imaging as Implemented at a Single-Site Major Referral Center.

MR imaging has become the routine technique for staging nasopharyngeal carcinoma, evaluating perineural tumor spread, and detecting cartilage invasion in laryngeal carcinoma. However, these protocols traditionally require in the range of 25 to 35 minutes of acquisition time. 3D sequences offer the potential advantage of time savings through the acquisition of 1-mm or submillimeter resolution isotropic data followed by multiplanar reformats that require no further imaging time. We have iteratively optimized vendor product 3D T1-weighted MR imaging sequences for morphologic face and neck imaging, reducing the average acquisition time of our 3T protocols by 9 minutes 57 seconds (40.9%) and of our 1.5T protocols by 9 minutes 5 seconds (37.0%), while simultaneously maintaining or improving spatial resolution. This clinical report describes our experience optimizing and implementing commercially available 3D T1-weighted MR imaging pulse sequence protocols for clinical face and neck MR imaging examinations using illustrative cases. We provide protocol details to allow others to replicate our implementations, and we report challenges we faced along with our solutions.

Patterns of Signal Intensity in CISS MRI of the Inner Ear and Eye.

Constructive interference in steady state (CISS) is a gradient echo magnetic resonance imaging (MRI) pulse sequence that provides excellent contrast between cerebrospinal fluid and adjacent structures but is prone to banding artifacts due to magnetic field inhomogeneities. We aimed to characterize artifacts in the inner ear and eye. In 30 patients (60 ears/eyes) undergoing CISS sequence MRI, nine low-signal intensity regions were identified in the inner ear and compared to temporal bone histopathology. The number and angle of bands across the eye were examined. In the cochlea, all ears had regions of low signal corresponding to anatomy (modiolus (all), spiral lamina (n = 59, 98.3%), and interscalar septa (n = 50, 83.3%)). In the labyrinth, the lateral semicircular canal crista (n = 42, 70%) and utricular macula (n = 47, 78.3%) were seen. Areas of low signal in the vestibule seen in all ears may represent the walls of the membranous utricle. Zero to three banding artifacts were seen in both eyes (right: 96.7%, mean 1.5; left: 93.3%, mean 1.3). Low signal regions in the inner ear on CISS sequences are common and have consistent patterns; most in the inner ear represent anatomy, appearing blurred due to partial volume averaging. Banding artifacts in the eye are more variable.

Development and validation of a multi-modality fusion deep learning model for differentiating glioblastoma from solitary brain metastases.

Glioblastoma (GBM) and brain metastases (BMs) are the two most common malignant brain tumors in adults. Magnetic resonance imaging (MRI) is a commonly used method for screening and evaluating the prognosis of brain tumors, but the specificity and sensitivity of conventional MRI sequences in differential diagnosis of GBM and BMs are limited. In recent years, deep neural network has shown great potential in the realization of diagnostic classification and the establishment of clinical decision support system. This study aims to apply the radiomics features extracted by deep learning techniques to explore the feasibility of accurate preoperative classification for newly diagnosed GBM and solitary brain metastases (SBMs), and to further explore the impact of multimodality data fusion on classification tasks. Standard protocol cranial MRI sequence data from 135 newly diagnosed GBM patients and 73 patients with SBMs confirmed by histopathologic or clinical diagnosis were retrospectively analyzed. First, structural T1-weight, T1C-weight, and T2-weight were selected as 3 inputs to the entire model, regions of interest (ROIs) were manually delineated on the registered three modal MR images, and multimodality radiomics features were obtained, dimensions were reduced using a random forest (RF)-based feature selection method, and the importance of each feature was further analyzed. Secondly, we used the method of contrast disentangled to find the shared features and complementary features between different modal features. Finally, the response of each sample to GBM and SBMs was predicted by fusing 2 features from different modalities. The radiomics features using machine learning and the multi-modal fusion method had a good discriminatory ability for GBM and SBMs. Furthermore, compared with single-modal data, the multimodal fusion models using machine learning algorithms such as support vector machine (SVM), Logistic regression, RF, adaptive boosting (AdaBoost), and gradient boosting decision tree (GBDT) achieved significant improvements, with area under the curve (AUC) values of 0.974, 0.978, 0.943, 0.938, and 0.947, respectively; our comparative disentangled multi-modal MR fusion method performs well, and the results of AUC, accuracy (ACC), sensitivity (SEN) and specificity(SPE) in the test set were 0.985, 0.984, 0.900, and 0.990, respectively. Compared with other multi-modal fusion methods, AUC, ACC, and SEN in this study all achieved the best performance. In the ablation experiment to verify the effects of each module component in this study, AUC, ACC, and SEN increased by 1.6%, 10.9% and 15.0%, respectively after 3 loss functions were used simultaneously. A deep learning-based contrast disentangled multi-modal MR radiomics feature fusion technique helps to improve GBM and SBMs classification accuracy.