Set Of MR Images Research Articles

Gadoxetic acid-enhanced MRI for the detection of liver metastases from melanoma.

We aimed to assess imaging findings and detection sensitivity for melanoma liver metastases on gadoxetic acid-enhanced magnetic resonance imaging (MRI). This retrospective study included patients with melanoma liver metastasis who underwent gadoxetic acid-enhanced MRI. Two abdominal radiologists independently evaluated signal characteristics of liver metastases on morphologic imaging (precontrast T1- and T2-weighted imaging), diffusion-weighted imaging (DWI), dynamic imaging, and hepatobiliary phase (HBP). Imaging findings were compared according to detection on HBP and the primary site of the melanoma using logistic regression with the generalized estimating equation (GEE). Detection sensitivity for metastases was compared among different MR imaging sets using logistic regression with GEE. A total of 67 patients with 254 liver metastases were included (44 women; mean age ± standard deviation, 65.6 ± 13.0 years). On HBP, 76.0% of metastases were detected, and 55.5% (141/254) showed hypointensity. Most of the metastases that were not detected on HBP originated from ocular melanomas (98.4%, 60/61), ≤1 cm (90.2%, 55/61) and showed T1 hyperintensity (98.4%, 60/61). Metastases from non-ocular melanomas more frequently showed T1 hypointensity, T2 hyperintensity, diffusion restriction, arterial enhancement, and HBP hypointensity than those from ocular melanomas (Ps ≤ 0.019). The detection sensitivity of HBP (76.0%) was significantly higher than DWI (65.7%, P = 0.006), but lower than morphologic imaging (98.8%, P < 0.001) and dynamic imaging (97.6%, P < 0.001). The detection sensitivity of HBP for melanoma liver metastasis was 76.0%, which was lower than that of morphologic or dynamic imaging. HBP of gadoxetic acid-enhanced MRI has little advantage in detecting melanoma liver metastases.

Improving Diagnostic Performance of MRI for Temporal Lobe Epilepsy With Deep Learning-Based Image Reconstruction in Patients With Suspected Focal Epilepsy.

To evaluate the diagnostic performance and image quality of 1.5-mm slice thickness MRI with deep learning-based image reconstruction (1.5-mm MRI + DLR) compared to routine 3-mm slice thickness MRI (routine MRI) and 1.5-mm slice thickness MRI without DLR (1.5-mm MRI without DLR) for evaluating temporal lobe epilepsy (TLE). This retrospective study included 117 MR image sets comprising 1.5-mm MRI + DLR, 1.5-mm MRI without DLR, and routine MRI from 117 consecutive patients (mean age, 41 years; 61 female; 34 patients with TLE and 83 without TLE). Two neuroradiologists evaluated the presence of hippocampal or temporal lobe lesions, volume loss, signal abnormalities, loss of internal structure of the hippocampus, and lesion conspicuity in the temporal lobe. Reference standards for TLE were independently constructed by neurologists using clinical and radiological findings. Subjective image quality, signal-to-noise ratio (SNR), and contrast-to-noise ratio (CNR) were analyzed. Performance in diagnosing TLE, lesion findings, and image quality were compared among the three protocols. The pooled sensitivity of 1.5-mm MRI + DLR (91.2%) for diagnosing TLE was higher than that of routine MRI (72.1%, P < 0.001). In the subgroup analysis, 1.5-mm MRI + DLR showed higher sensitivity for hippocampal lesions than routine MRI (92.7% vs. 75.0%, P = 0.001), with improved depiction of hippocampal T2 high signal intensity change (P = 0.016) and loss of internal structure (P < 0.001). However, the pooled specificity of 1.5-mm MRI + DLR (76.5%) was lower than that of routine MRI (89.2%, P = 0.004). Compared with 1.5-mm MRI without DLR, 1.5-mm MRI + DLR resulted in significantly improved pooled accuracy (91.2% vs. 73.1%, P = 0.010), image quality, SNR, and CNR (all, P < 0.001). The use of 1.5-mm MRI + DLR enhanced the performance of MRI in diagnosing TLE, particularly in hippocampal evaluation, because of improved depiction of hippocampal abnormalities and enhanced image quality.

A Comparison of Treatment Plan Adaptation Methods on PTV Margin Coverage in MR-Linac Prostate SBRT.

MRI-Linac guided radiotherapy is a novel treatment modality allowing for daily online adaptation of contours and treatment planning. MRI-Linac prostate SBRT (MRL-SBRT) has recently been shown to provide a tangible benefit in decreasing both acute and late toxicity, due to the ability to evaluate soft tissue contrast, directly monitor organ motion, and anatomical changes during the entire treatment course. Multiple MRI-Linac platforms are available, which provide different advantages with regards to imaging and motion management. We undertook this study to assess coverage of the clinical target volume with use of a standard PTV margin in patients receiving either adapt-to-position (ATP) only treatment or a combination of adapt-to-position and adapt-to-shape (ATP+ATS) on the Elekta Unity MR-Linac. Data was collected retrospectively in patients who underwent prostate SBRT using a 1.5 T MR-Linac (Unity, Elekta AB, Stockholm, Sweden). Prior to each fraction, determination of ATP or ATS planning was at the discretion of the treating physician. Prostate and organs at risk (OAR's) were contoured on the reference MRI and daily MR image sets acquired prior to each fraction. The prostate PTV consisted of a 5mm margin, except 3mm posteriorly. Contours were merged to account for target and OAR volumes during all fractions. Target volumes from the reference plan were then compared to the clinical target volume (CTV) on pre-treatment MR-Linac imaging as well as intersection with OAR's. Nineteen men, median age 72 years (range 67-75), with prostate cancer underwent treatment from May 2021 to January 2023 with MRL-SBRT. The average time for ATP vs ATP+ATS was 45 minutes (range 35-60) and 60 minutes (range 50-80), respectively. The mean CTV treated was 43.4cc (range 31-60) for ATP and 51.8cc (range 36-67) for ATP+ATS. Prior to plan adaptation, target volumes in 92.6% of fractions were not covered completely by the PTV. In the ATP cohort, 89% of fractions had a portion of the CTV uncovered with an average of 2.9cc. In the ATP+ATS cohort, 96.4% had a portion of the CTV uncovered, with an average of 6.9cc. There was no difference in the amount of PTV intersection for ATP or ATP+ATS plans with the rectum (1.1cc vs. 1.5cc) and bladder (9.2cc vs. 12.6cc), respectively. All patients completed the planned course of treatment. Adaptive radiation therapy for prostate cancer utilizing an MR-Linac based approach with the Elekta Unity was feasible for patients requiring either ATP only or ATP+ATS plans. As expected, ATS plans required longer treatment time, allowing for improved coverage of the target when deemed clinically necessary. Despite a larger overall volume treated in ATS plans, there was no difference in overlap with OARs between ATP and ATS plans. Adaption for each daily fraction and intrafraction motion monitoring improved CT coverage with our pre-specified PTV margin.

Virtual 4DCT generated from 4DMRI in gated particle therapy: phantom validation and application to lung cancer patients

Objective. Respiration negatively affects the outcome of a radiation therapy treatment, with potentially severe effects especially in particle therapy (PT). If compensation strategies are not applied, accuracy cannot be achieved. To support the clinical practice based on 4D computed tomography (CT), 4D magnetic resonance imaging (MRI) acquisitions can be exploited. The purpose of this study was to validate a method for virtual 4DCT generation from 4DMRI data for lung cancers on a porcine lung phantom, and to apply it to lung cancer patients in PT. Approach. Deformable image registration was used to register each respiratory phase of the 4DMRI to a reference phase. Then, a static 3DCT was registered to this reference MR image set, and the virtual 4DCT was generated by warping the registered CT according to previously obtained deformation fields. The method was validated on a physical phantom for which a ground truth 4DCT was available and tested on lung tumor patients, treated with gated PT at end-exhale, by comparing the virtual 4DCT with a re-evaluation 4DCT. The geometric and dosimetric evaluation was performed for both proton and carbon ion treatment plans. Main results. The phantom validation exhibited a geometrical accuracy within the maximum resolution of the MRI and mean dose deviations, with respect to the prescription dose, up to 3.2% for target D 95%, with a mean gamma pass rate of 98%. For patients, the virtual and re-evaluation 4DCTs showed good correspondence, with errors on target D 95% up to 2% within the gating window. For one patient, dose variations up to 10% at end-exhale were observed due to relevant inter-fraction anatomo-pathological changes that occurred between the planning and re-evaluation CTs. Significance. Results obtained on phantom data showed that the virtual 4DCT method was accurate, allowing its application on patient data for testing within a clinical scenario.

PigSNIPE: Scalable Neuroimaging Processing Engine for Minipig MRI

Translation of basic animal research to find effective methods of diagnosing and treating human neurological disorders requires parallel analysis infrastructures. Small animals such as mice provide exploratory animal disease models. However, many interventions developed using small animal models fail to translate to human use due to physical or biological differences. Recently, large-animal minipigs have emerged in neuroscience due to both their brain similarity and economic advantages. Medical image processing is a crucial part of research, as it allows researchers to monitor their experiments and understand disease development. By pairing four reinforcement learning models and five deep learning UNet segmentation models with existing algorithms, we developed PigSNIPE, a pipeline for the automated handling, processing, and analyzing of large-scale data sets of minipig MR images. PigSNIPE allows for image registration, AC-PC alignment, detection of 19 anatomical landmarks, skull stripping, brainmask and intracranial volume segmentation (DICE 0.98), tissue segmentation (DICE 0.82), and caudate-putamen brain segmentation (DICE 0.8) in under two minutes. To the best of our knowledge, this is the first automated pipeline tool aimed at large animal images, which can significantly reduce the time and resources needed for analyzing minipig neuroimages.

Impact of MRI resolution for Linac-based stereotactic radiosurgery.

Magnetic resonance imaging (MRI) is a standard imaging modality in intracranial stereotactic radiosurgery (SRS) for defining target volumes. However, wide disparities in MRI resolution exist, which could directly impact accuracy of target delineation. Here, sequences with various MRI resolution were acquired on phantoms to evaluate the effect on volume definition and dosimetric consequence for cranial SRS. Four T1-weighted MR sequences with increasing 3D resolution were compared, including two Spin Echo (SE) 2D acquisitions with 5mm and 3mm slice thickness (SE5mm, SE3mm) and two gradient echo 3D acquisitions (TFE, BRAVO). The voxel sizes were 0.4×0.4×5.0, 0.5×0.5×3.0, 0.9×0.9×1.25, and 0.4×0.4×0.5 mm3, respectively. Four phantoms with simulated lesions of different shape and volume (range, 0.53-25.0 cm3) were imaged, resulting in 16 total sets of MRIs. Four radiation oncologists provided contours on individual MR image set. All observer contours were compared with ground truth, defined on CT image according to the absolute dimensions of the target structure, using Dice similarity coefficient (DSC), Hausdorff distance (HD), mean distance-to-agreement (MDA), and the ratio between reconstructed and true volume (Ratiovol ). For dosimetric consequence, SRS plans targeting observer volumes were created. The true Paddick conformity index ( ), calculated with true target volume, was correlated with quality of observer volume. All measures of observer contours improved as increasingly higher MRI resolution was provided from SE5mm to BRAVO. The improvement in DSC, HD and MDA was statistically significant (p<0.01). Dosimetrically, strongly correlated with DSC of the planning observer volume (Pearson's r=0.94, p<0.00001). Significant improvement in target definition and reduced inter-observer variation was observed as the MRI resolution improved, which also improved the quality of SRS plans. Results imply that high resolution 3D MR sequences should be used to minimize potential errors in target definition, and multi-slice 2D sequences should be avoided.

Synthetic MR image generation of macrotrabecular-massive hepatocellular carcinoma using generative adversarial networks

PurposeThe purpose of this study was to develop a method for generating synthetic MR images of macrotrabecular-massive hepatocellular carcinoma (MTM-HCC). Materials and methodsA set of abdominal MR images including fat-saturated T1-weighted images obtained during the arterial and portal venous phases of enhancement and T2-weighted images of 91 patients with MTM-HCC, and another set of MR abdominal images from 67 other patients were used. Synthetic images were obtained using a 3-step pipeline that consisted in: (i), generating a synthetic MTM-HCC tumor on a neutral background; (ii), randomly selecting a background among the 67 patients and a position inside the liver; and (iii), merging the generated tumor in the background at the specified location. Synthetic images were qualitatively evaluated by three radiologists and quantitatively assessed using a mix of 1-nearest neighbor classifier metric and Fréchet inception distance. ResultsA set of 1000 triplets of synthetic MTM-HCC images with consistent contrasts were successfully generated. Evaluation of selected synthetic images by three radiologists showed that the method gave realistic, consistent and diversified images. Qualitative and quantitative evaluation led to an overall score of 0.64. ConclusionThis study shows the feasibility of generating realistic synthetic MR images with very few training data, by leveraging the wide availability of liver backgrounds. Further studies are needed to assess the added value of those synthetic images for automatic diagnosis of MTM-HCC.

A novel data-driven method for the analysis and reconstruction of cardiac cine MRI

Cardiac cine magnetic resonance imaging (MRI) can be considered the optimal criterion for measuring cardiac function. This imaging technique can provide us with detailed information about cardiac structure, tissue composition and even blood flow, which makes it highly used in medical science. But due to the image time acquisition and several other factors the MRI sequences can easily get corrupted, causing radiologists to misdiagnose 40 million people worldwide each and every single year. Hence, the urge to decrease these numbers, researchers from different fields have been introducing novel tools and methods in the medical field. Aiming to the same target, we consider in this work the application of the higher order dynamic mode decomposition (HODMD) technique. The HODMD algorithm is a linear method, which was originally introduced in the fluid dynamics domain, for the analysis of complex systems. Nevertheless, the proposed method has extended its applicability to numerous domains, including medicine. In this work, HODMD in used to analyze sets of MR images of a heart, with the ultimate goal of identifying the main patterns and frequencies driving the heart dynamics. Furthermore, a novel interpolation algorithm based on singular value decomposition combined with HODMD is introduced, providing a three-dimensional reconstruction of the heart. This algorithm is applied (i) to reconstruct corrupted or missing images, and (ii) to build a reduced order model of the heart dynamics.

High-Resolution Three-Dimensional Hybrid MRI + Low Dose CT Vocal Tract Modeling: A Cadaveric Pilot Study

<h2>Summary</h2><h3>Objectives</h3> MRI based vocal tract models have many applications in voice research and education. These models do not adequately capture bony structures (e.g. teeth, mandible), and spatial resolution is often relatively low in order to minimize scanning time. Most MRI sequences achieve 3D vocal tract coverage at gross resolutions of 2 mm³ within a scan time of <20 seconds. Computed tomography (CT) is well suited for vocal tract imaging, but is infrequently used due to the risk of ionizing radiation. In this cadaveric study, a single, extremely low-dose CT scan of the bony structures is blended with accelerated high-resolution (1 mm³) MRI scans of the soft tissues, creating a high-resolution hybrid CT-MRI vocal tract model. <h3>Methods</h3> Minimum CT dosages were determined and a custom 16-channel airway receiver coil for accelerated high (1 mm³) resolution MRI was evaluated. A rigid body landmark based partial volume registration scheme was then applied to the images, creating a hybrid CT-MRI model that was segmented in Slicer. <h3>Results</h3> Ultra-low dose CT produced images with sufficient quality to clearly visualize the bone, and exposed the cadaver to 0.06 mSv. This is comparable to atmospheric exposures during a round trip transatlantic flight. The custom 16-channel vocal tract coil produced acceptable image quality at 1 mm³ resolution when reconstructed from ∼6 fold undersampled data. High (1 mm³) resolution MR imaging of short (<10 seconds) sustained sounds was achieved. The feasibility of hybrid CT-MRI vocal tract modeling was successfully demonstrated using the rigid body landmark based partial volume registration scheme. Segmentations of CT and hybrid CT-MRI images provided more detailed 3D representations of the vocal tract than 2 mm³ MRI based segmentations. <h3>Conclusions</h3> The method described in this study indicates that high-resolution CT and MR image sets can be combined so that structures such as teeth and bone are accurately represented in vocal tract reconstructions. Such scans will aid learning and deepen understanding of anatomical features that relate to voice production, as well as furthering knowledge of the static and dynamic functioning of individual structures relating to voice production.

Quantification of systemic-to-pulmonary collateral flow in univentricular physiology with 4D flow MRI.

Systemic-to-pulmonary collateral flow is a well-recognised phenomenon in patients with single ventricle physiology, but remains difficult to quantify. The aim was to compare the reported formula's that have been used for calculation of systemic-to-pulmonary-collateral flow to assess their consistency and to quantify systemic-to-pulmonary collateral flow in patients with a Glenn and/or Fontan circulation using four-dimensional flow MRI (4D flow MR). Retrospective case-control study of Glenn and Fontan patients who had a 4D flow MR study. Flows were measured at the ascending aorta, left and right pulmonary arteries, left and right pulmonary veins, and both caval veins. Systemic-to-pulmonary collateral flow was calculated using two formulas: 1) pulmonary veins - pulmonary arteries and 2) ascending aorta - caval veins. Anatomical identification of collaterals was performed using the 4D MR image set. Fourteen patients (n = 11 Fontan, n = 3 Glenn) were included (age 26 [22-30] years). Systemic-to-pulmonary collateral flow was significantly higher in the patients than the controls (n = 10, age 31.2 [15.1-38.4] years) with both formulas: 0.28 [0.09-0.5] versus 0.04 [-0.66-0.21] l/min/m2 (p = 0.036, formula 1) and 0.67 [0.24-0.88] versus -0.07 [-0.16-0.08] l/min/m2 (p < 0.001, formula 2). In patients, systemic-to-pulmonary collateral flow differed significantly between formulas 1 and 2 (13% versus 26% of aortic flow, p = 0.038). In seven patients, veno-venous collaterals were detected and no aortopulmonary collaterals were visualised. 4D flow MR is able to detect increased systemic-to-pulmonary collateral flow and visualise collaterals vessels in Glenn and Fontan patients. However, the amount of systemic-to-pulmonary collateral flow varies with the formula employed. Therefore, further research is necessary before it could be applied in clinical care.

Automated knee cartilage segmentation for heterogeneous clinical MRI using generative adversarial networks with transfer learning.

This study aimed to build a deep learning model to automatically segment heterogeneous clinical MRI scans by optimizing a pre-trained model built from a homogeneous research dataset with transfer learning. Conditional generative adversarial networks pretrained on the Osteoarthritis Initiative MR images was transferred to 30 sets of heterogenous MR images collected from clinical routines. Two trained radiologists manually segmented the 30 sets of clinical MR images for model training, validation and test. The model performance was compared to models trained from scratch with different datasets, as well as two radiologists. A 5-fold cross validation was performed. The transfer learning model obtained an overall averaged Dice coefficient of 0.819, an averaged 95 percentile Hausdorff distance of 1.463 mm, and an averaged average symmetric surface distance of 0.350 mm on the 5 random holdout test sets. A 5-fold cross validation had a mean Dice coefficient of 0.801, mean 95 percentile Hausdorff distance of 1.746 mm, and mean average symmetric surface distance of 0.364 mm. It outperformed other models and performed similarly as the radiologists. A transfer learning model was able to automatically segment knee cartilage, with performance comparable to human, using heterogeneous clinical MR images with a small training data size. In addition, the model proved robust when tested through cross validation and on images from a different vendor. We found it feasible to perform fully automated cartilage segmentation of clinical knee MR images, which would facilitate the clinical application of quantitative MRI techniques and other prediction models for improved patient treatment planning.

Evaluation of Multisource Adaptive MRI Fusion for Gross Tumor Volume Delineation of Hepatocellular Carcinoma.

PurposeTumor delineation plays a critical role in radiotherapy for hepatocellular carcinoma (HCC) patients. The incorporation of MRI might improve the ability to correctly identify tumor boundaries and delineation consistency. In this study, we evaluated a novel Multisource Adaptive MRI Fusion (MAMF) method in HCC patients for tumor delineation.MethodsTen patients with HCC were included in this study retrospectively. Contrast-enhanced T1-weighted MRI at portal-venous phase (T1WPP), contrast-enhanced T1-weighted MRI at 19-min delayed phase (T1WDP), T2-weighted (T2W), and diffusion-weighted MRI (DWI) were acquired on a 3T MRI scanner and imported to in-house-developed MAMF software to generate synthetic MR fusion images. The original multi-contrast MR image sets were registered to planning CT by deformable image registration (DIR) using MIM. Four observers independently delineated gross tumor volumes (GTVs) on the planning CT, four original MR image sets, and the fused MRI for all patients. Tumor contrast-to-noise ratio (CNR) and Dice similarity coefficient (DSC) of the GTVs between each observer and a reference observer were measured on the six image sets. Inter-observer and inter-patient mean, SD, and coefficient of variation (CV) of the DSC were evaluated.ResultsFused MRI showed the highest tumor CNR compared to planning CT and original MR sets in the ten patients. The mean ± SD tumor CNR was 0.72 ± 0.73, 3.66 ± 2.96, 4.13 ± 3.98, 4.10 ± 3.17, 5.25 ± 2.44, and 9.82 ± 4.19 for CT, T1WPP, T2W, DWI, T1WDP, and fused MRI, respectively. Fused MRI has the minimum inter-observer and inter-patient variations as compared to original MR sets and planning CT sets. GTV delineation inter-observer mean DSC across the ten patients was 0.81 ± 0.09, 0.85 ± 0.08, 0.88 ± 0.04, 0.89 ± 0.08, 0.90 ± 0.04, and 0.95 ± 0.02 for planning CT, T1WPP, T2W, DWI, T1WDP, and fused MRI, respectively. The patient mean inter-observer CV of DSC was 3.3%, 3.2%, 1.7%, 2.6%, 1.5%, and 0.9% for planning CT, T1WPP, T2W, DWI, T1WDP, and fused MRI, respectively.ConclusionThe results demonstrated that the fused MRI generated using the MAMF method can enhance tumor CNR and improve inter-observer consistency of GTV delineation in HCC as compared to planning CT and four commonly used MR image sets (T1WPP, T1WDP, T2W, and DWI). The MAMF method holds great promise in MRI applications in HCC radiotherapy treatment planning.

Brain Tumor Classification Using a Combination of Variational Autoencoders and Generative Adversarial Networks.

Brain tumors are a pernicious cancer with one of the lowest five-year survival rates. Neurologists often use magnetic resonance imaging (MRI) to diagnose the type of brain tumor. Automated computer-assisted tools can help them speed up the diagnosis process and reduce the burden on the health care systems. Recent advances in deep learning for medical imaging have shown remarkable results, especially in the automatic and instant diagnosis of various cancers. However, we need a large amount of data (images) to train the deep learning models in order to obtain good results. Large public datasets are rare in medicine. This paper proposes a framework based on unsupervised deep generative neural networks to solve this limitation. We combine two generative models in the proposed framework: variational autoencoders (VAEs) and generative adversarial networks (GANs). We swap the encoder–decoder network after initially training it on the training set of available MR images. The output of this swapped network is a noise vector that has information of the image manifold, and the cascaded generative adversarial network samples the input from this informative noise vector instead of random Gaussian noise. The proposed method helps the GAN to avoid mode collapse and generate realistic-looking brain tumor magnetic resonance images. These artificially generated images could solve the limitation of small medical datasets up to a reasonable extent and help the deep learning models perform acceptably. We used the ResNet50 as a classifier, and the artificially generated brain tumor images are used to augment the real and available images during the classifier training. We compared the classification results with several existing studies and state-of-the-art machine learning models. Our proposed methodology noticeably achieved better results. By using brain tumor images generated artificially by our proposed method, the classification average accuracy improved from 72.63% to 96.25%. For the most severe class of brain tumor, glioma, we achieved 0.769, 0.837, 0.833, and 0.80 values for recall, specificity, precision, and F1-score, respectively. The proposed generative model framework could be used to generate medical images in any domain, including PET (positron emission tomography) and MRI scans of various parts of the body, and the results show that it could be a useful clinical tool for medical experts.

Mass Sphering Approach for Three-Dimensional Reconstruction of Brain Tumor using MRI

Objective: The objective of the research work is to reconstruct the brain tumor three-dimensionally with high degree of accuracy. Methods: This study describes 3D reconstruction of brain tumor using Mass Sphering Approach (MSA) algorithm. 39 weighted features are extracted from the non-tumor and tumor pixels. These weighted features are used to train the Support Vector Machine (SVM) algorithm. Number of training samples taken to train SVM algorithm are 268 and the testing sample are 64. The complete MR image set of a subject (64 axial slices) are detected for tumor pixels and these slices are concatenated to get volumetric tumor information. Findings: 5-step MSA algorithm is proposed which includes filtering, segmentation, classification, optimization and reconstruction. MR images are subjected to Rician noise which can be removed by a simple correction scheme, initiated to change the bias due to the Rician distribution of the noisy magnitude data. The filtered MR image slices are segmented and classified to detect the tumor areas and the tumor pixels are subjected for 3D reconstruction. The improvement in performance of MSA is depicted by comparing the algorithm with traditional SVM. Novelty: The accuracy achieved in detecting glioma and glioblastoma using MSA are 95.24% and 99% respectively which is highly remarkable. Keywords: Glioma; voxel; Magnetic Resonance; Classification; Immune; Reconstruction

Fuzzy Inference Based FMEA for MR Image Based HDR Brachytherapy

Purpose/Objective(s)MR image based HDR brachytherapy is a complex, time-constrained and resource-intensive procedure requiring significant coordination and cooperation of multi-departmental efforts. Failure modes and effects analysis (FMEA) is an effective risk evaluation tool for identifying all possible failure modes (FMs) that could potentially impact negatively on the patient's treatment. Given the complexity of the MR image based HDR brachytherapy clinical workflow, FMEA can play a critical role in improving quality measures for each task. However, assigning a value between 1 to 10 to each of risk factor in traditional FMEA is a subjective and challenging task for the relevant multi-disciplinary team and potentially can cause considerable discrepancy and inconsistency. In this study, we introduced fuzzy inference based FMEA to provide better understanding of FMs to improve the quality assurance program for an efficient workflow and improved patient safety.Materials/MethodsA total of four major sub-processes and forty-five FMs were identified from “applicator insertion” to “treatment delivery”. The fuzzy inference rule-based model uses three inputs of risk factors (severity, occurrence, and detectability) and single output of fuzzy risk priority number (RPN) consisting of five fuzzy linguistic terms (very low, low, medium, high, and very high) of triangular membership functions. This rule-based model is generated based on a set of fuzzy if-then rules by decision makers that relate input to output variables. Then, the defuzzification method was used to obtain a fuzzy RPN value.ResultsThe team identified top three FMs of traditional FMEA. These were incorrect insertion of applicators by wrong identification ultrasound track, applicator moves, and incorrect channel length. Different rankings were produced depending on the rules of decision maker in fuzzy inference FMEA. Top 3 FMs of fuzzy FMEA were identified as missed/not detecting errors during physics second check, wrong applicator reconstruction, and poor quality of MR image set. The differences between the two RPN rankings are due to rules of decision making considering the probability of detecting errors in later process and severity.ConclusionThis is the first report for fuzzy inference based FMEA on MRI based HDR brachytherapy. The fuzzy FMEA model exhibits desirable properties that help overcome the drawbacks of the traditional FMEA and RPN, especially when dealing with multi-disciplinary team practice where their priorities on FMs are different. The results have been shared with team members, and mitigation plans are to be implemented for enhancing quality of care and improving patient safety. MR image based HDR brachytherapy is a complex, time-constrained and resource-intensive procedure requiring significant coordination and cooperation of multi-departmental efforts. Failure modes and effects analysis (FMEA) is an effective risk evaluation tool for identifying all possible failure modes (FMs) that could potentially impact negatively on the patient's treatment. Given the complexity of the MR image based HDR brachytherapy clinical workflow, FMEA can play a critical role in improving quality measures for each task. However, assigning a value between 1 to 10 to each of risk factor in traditional FMEA is a subjective and challenging task for the relevant multi-disciplinary team and potentially can cause considerable discrepancy and inconsistency. In this study, we introduced fuzzy inference based FMEA to provide better understanding of FMs to improve the quality assurance program for an efficient workflow and improved patient safety. A total of four major sub-processes and forty-five FMs were identified from “applicator insertion” to “treatment delivery”. The fuzzy inference rule-based model uses three inputs of risk factors (severity, occurrence, and detectability) and single output of fuzzy risk priority number (RPN) consisting of five fuzzy linguistic terms (very low, low, medium, high, and very high) of triangular membership functions. This rule-based model is generated based on a set of fuzzy if-then rules by decision makers that relate input to output variables. Then, the defuzzification method was used to obtain a fuzzy RPN value. The team identified top three FMs of traditional FMEA. These were incorrect insertion of applicators by wrong identification ultrasound track, applicator moves, and incorrect channel length. Different rankings were produced depending on the rules of decision maker in fuzzy inference FMEA. Top 3 FMs of fuzzy FMEA were identified as missed/not detecting errors during physics second check, wrong applicator reconstruction, and poor quality of MR image set. The differences between the two RPN rankings are due to rules of decision making considering the probability of detecting errors in later process and severity. This is the first report for fuzzy inference based FMEA on MRI based HDR brachytherapy. The fuzzy FMEA model exhibits desirable properties that help overcome the drawbacks of the traditional FMEA and RPN, especially when dealing with multi-disciplinary team practice where their priorities on FMs are different. The results have been shared with team members, and mitigation plans are to be implemented for enhancing quality of care and improving patient safety.

Evaluation of cortical superficial siderosis in patients with cognitive dysfunction using 3D FLAIR and 3D DIR.

Our aim was to evaluate the detectability of cortical superficial siderosis (cSS) by 3D FLAIR and 3D DIR images in comparison with the SWI images in patients with cognitive dysfunction. We studied 246 patients with cognitive dysfunction (144 women, 102 men; mean age: 75.5 ± 7.53years) who visited a memory clinic at our hospital and underwent MR examinations at 3T. Specifically, 16 patients with Alzheimer disease (AD) (n = 11) and AD with cerebrovascular disease (n = 5) manifested cSS based on SWI. Each set of MR images (3D FLAIR and 3D DIR) was reviewed by two reviewers separately for the detection of sulcal hyperintensity that suggested cSS. SWI detected a greater number of cSS sulci than 3D DIR and 3D FLAIR. The sensitivity and specificity for the detection of sulcal hyperintensity were the same between 3D FLAIR and 3D DIR (87.5%/100%). However, 3D DIR detected a greater number of cSS sulci than 3D FLAIR (p = .005). Our study showed that 3D DIR and 3D FLAIR can detect sulcal hyperintensity related to cSS although they are less sensitive to cSS lesions than SWI. • 3D FLAIR and 3D DIR can show sulcal signal abnormalities related to cSS in patients with cognitive dysfunction. • 3D FLAIR and 3D DIR detect sulcal hyperintensity of cSS, although they are less sensitive to cSS than SWI. • Signal alterations due to cSS are more detectable in 3D DIR than in 3D FLAIR.

A phantom study comparing radial trajectories for accelerated cardiac 4D flow MRI against a particle imaging velocimetry reference.

Radial sampling is one method to accelerate 4D flow MRI acquisition, making feasible dual-velocity encoding (Venc) assessment of slow flow in the left ventricle (LV). Here, two radial trajectories are compared in vitro for this application: 3D radial (phase-contrast vastly undersampled isotropic projection, PC-VIPR) versus stack of stars (phase-contrast stack of stars, PC-SOS), with benchtop particle imaging velocimetry (PIV) serving as a reference standard. The study contained three steps: (1) Construction of an MRI- and PIV-compatible LV model from a healthy adult's CT images. (2) In vitro PIV using a pulsatile flow pump. (3) In vitro dual-Venc 4D flow MRI using PC-VIPR and PC-SOS (two repeat experiments). Each MR image set was retrospectively undersampled to five effective scan durations and compared with the PIV reference. The root-mean-square velocity vector difference (RMSE) between MRI and PIV images was compared, along with kinetic energy (KE) and wall shear stress (WSS). RMSE increased as scan time decreased for both MR acquisitions. RMSE was 3% lower in PC-SOS images than PC-VIPR images in 30-min scans (3.8 vs. 3.9 cm/s) but 98% higher in 2.5-min scans (9.5 vs. 4.8 cm/s). PIV intrasession repeatability showed a RMSE of 4.4 cm/s, reflecting beat-to-beat flow variation, while MRI had intersession RMSEs of 3.8/3.5 cm/s for VIPR/SOS, respectively. Speed, KE, and WSS were overestimated voxel-wise in 30-min MRI scans relative to PIV by 0.4/0.3 cm/s, 0.2/0.1 μJ/mL, and 36/43 mPa, respectively, for VIPR/SOS. PIV is feasible for application-specific 4D flow MRI protocol optimization. PC-VIPR is better-suited to dual-Venc LV imaging with short scan times.

Multimodal Medical Image Fusion Based on Gabor Representation Combination of Multi-CNN and Fuzzy Neural Network

Aiming at the current multimodal medical image fusion methods that cannot fully characterize the complex textures and edge information of the lesion in the fused image, a method based on Gabor representation of multi-CNN combination and fuzzy neural network is proposed. This method first filters the CT and MR image sets through a set of Gabor filter banks with different proportions and directions to obtain different Gabor representations pairs of CT and MR, each pair of different Gabor representations is used to train the corresponding CNN to generate a G- CNN and multiple G- CNN form a G- CNN group, namely G- CNNs; then when fusing CT and MR images, CT and MR are represented by Gabors to get Gabor representation pairs firstly, each Gabor representation pair is put into the corresponding trained G- CNN for preliminary fusion, then use the fuzzy neural network to fuse multiple outputs of the G- CNNs to obtain the final fused image. Compared with the nine recent state-of-the-art multimodal fusion methods, the average mutual information of the three groups of experiments has increased by 13%, 10.3%, and 10% respectively; the average spatial frequency has increased by 10.3%, 20%, and 10.7%; the average standard deviation has increased respectively 12.4%, 10.8%, 14.4%; the average edge retention information increased by 33.5%, 22%, and 43%. The experimental results show that the proposed fusion method is significantly better than the other comparative fusion methods in objective evaluation and visual quality. It has the best performance on the four indicators and can better integrate the rich texture features and the clear edge information of the source images into the final fused image, which improves the quality of multimodal medical image fusion, and effectively assists doctors in disease diagnosis.

Can MRI Biomarkers Predict Triple-Negative Breast Cancer?

The purpose of this study was to investigate MRI features of triple-negative breast cancer (TNBC) compared with non-TNBC, to predict histopathological results. In the study, 26 patients with TNBC and 24 with non-TNBC who underwent multiparametric MRI of the breast on a 3 T magnet over a 10-months period were retrospectively recruited. MR imaging sets were evaluated by two experienced breast radiologists in consensus and classified according to the 2013 American College of Radiology (ACR) BI-RADS lexicon. The comparison between the two groups was performed using the Chi-square test and followed by logistic regression analyses. We found that 92% of tumors presented as mass enhancements (p = 0.192). 41.7% of TNBC and 86.4% of non-TNBC had irregular shape (p = 0.005); 58.3% of TNBC showed circumscribed margins, compared to 9.1% of non-TNBC masses (p = 0.001); 75% of TNBC and 9.1% of non-TNBC showed rim enhancement (p < 0.001). Intralesional necrosis was significantly associated with TNBC (p = 0.016). Rim enhancement and intralesional necrosis risulted to be positive predictors at univariate analysis (OR = 29.86, and 8.10, respectively) and the multivariate analysis confirmed that rim enhancement is independently associated with TNBC (OR = 33.08). The mean ADC values were significantly higher for TNBC (p = 0.011). In conclusion, TNBC is associated with specific MRI features that can be possible predictors of pathological results, with a consequent prognostic value.

Deep-learning-based segmentation of the vocal tract and articulators in real-time magnetic resonance images of speech

Background and ObjectiveMagnetic resonance (MR) imaging is increasingly used in studies of speech as it enables non-invasive visualisation of the vocal tract and articulators, thus providing information about their shape, size, motion and position. Extraction of this information for quantitative analysis is achieved using segmentation. Methods have been developed to segment the vocal tract, however, none of these also fully segment any articulators. The objective of this work was to develop a method to fully segment multiple groups of articulators as well as the vocal tract in two-dimensional MR images of speech, thus overcoming the limitations of existing methods. MethodsFive speech MR image sets (392 MR images in total), each of a different healthy adult volunteer, were used in this work. A fully convolutional network with an architecture similar to the original U-Net was developed to segment the following six regions in the image sets: the head, soft palate, jaw, tongue, vocal tract and tooth space. A five-fold cross-validation was performed to investigate the segmentation accuracy and generalisability of the network. The segmentation accuracy was assessed using standard overlap-based metrics (Dice coefficient and general Hausdorff distance) and a novel clinically relevant metric based on velopharyngeal closure. ResultsThe segmentations created by the method had a median Dice coefficient of 0.92 and a median general Hausdorff distance of 5mm. The method segmented the head most accurately (median Dice coefficient of 0.99), and the soft palate and tooth space least accurately (median Dice coefficients of 0.92 and 0.93 respectively). The segmentations created by the method correctly showed 90% (27 out of 30) of the velopharyngeal closures in the MR image sets. ConclusionsAn automatic method to fully segment multiple groups of articulators as well as the vocal tract in two-dimensional MR images of speech was successfully developed. The method is intended for use in clinical and non-clinical speech studies which involve quantitative analysis of the shape, size, motion and position of the vocal tract and articulators. In addition, a novel clinically relevant metric for assessing the accuracy of vocal tract and articulator segmentation methods was developed.