To develop a machine learning (ML) model that predicts disease groups or autoantibodies in patients with idiopathic inflammatory myopathies (IIMs) using muscle MRI radiomics features. Twenty-two ...patients with dermatomyositis (DM), 14 with amyopathic dermatomyositis (ADM), 19 with polymyositis (PM) and 19 with non-IIM were enrolled. Using 2D manual segmentation, 93 original features as well as 93 local binary pattern (LBP) features were extracted from MRI (short-tau inversion recovery STIR imaging) of proximal limb muscles. To construct and compare ML models that predict disease groups using each set of features, dimensional reductions were performed using a reproducibility analysis by inter-reader and intra-reader correlation coefficients, collinearity analysis, and the sequential feature selection (SFS) algorithm. Models were created using the linear discriminant analysis (LDA), quadratic discriminant analysis (QDA), support vector machine (SVM), k-nearest neighbors (k-NN), random forest (RF) and multi-layer perceptron (MLP) classifiers, and validated using tenfold cross-validation repeated 100 times. We also investigated whether it was possible to construct models predicting autoantibody status. Our ML-based MRI radiomics models showed the potential to distinguish between PM, DM, and ADM. Models using LBP features provided better results, with macro-average AUC values of 0.767 and 0.714, accuracy of 61.2 and 61.4%, and macro-average recall of 61.9 and 59.8%, in the LDA and k-NN classifiers, respectively. In contrast, the accuracies of radiomics models distinguishing between non-IIM and IIM disease groups were low. A subgroup analysis showed that classification models for anti-Jo-1 and anti-ARS antibodies provided AUC values of 0.646-0.853 and 0.692-0.792, with accuracy of 71.5-81.0 and 65.8-78.3%, respectively. ML-based TA of muscle MRI may be used to predict disease groups or the autoantibody status in patients with IIM and is useful in non-invasive assessments of disease mechanisms.
Purpose
To generate a new discrimination method to distinguish between malignant mesenchymal tumors of the uterus and T2-weighted hyperintense leiomyoma based on magnetic resonance imaging findings ...and clinical features.
Materials and methods
Data from 32 tumors of 32 patients with malignant mesenchymal tumors of the uterus and from 34 tumors of 30 patients with T2-weighted hyperintense leiomyoma were analyzed. Clinical parameters, qualitative magnetic resonance imaging features, including computed diffusion-weighted imaging, and quantitative characteristics of magnetic resonance imaging of these two tumor types were compared. Predictive values for malignant mesenchymal tumors of the uterus were calculated using variant discriminant analysis.
Results
The T1 bright area on qualitative assessment and mean apparent diffusion coefficient value on quantitative assessment yielded the most independent magnetic resonance imaging differentiators of malignant mesenchymal tumors of the uterus and T2-weighted hyperintense leiomyoma. The classification accuracy of the variant discriminant analysis based on three selected findings, i.e., a T1 bright area, computed diffusion-weighted imaging with a b-value of 2000s/mm
2
(cDWI
2000
), and T2-hypointense bands, was 84.8% (56/66), indicating high accuracy.
Conclusions
Variant discriminant analysis using the T1 bright area, cDWI
2000
, and T2-hypointense bands yielded high accuracy for differentiating between malignant mesenchymal tumors of the uterus and T2-weighted hyperintense leiomyoma.
Magnetic resonance imaging (MRI) is indispensable in clinical medicine for the morphological and tomographic evaluation of many parenchymal organs. With varied imaging methods, diverse biological ...information, such as the perfusion volume and measurements of metabolic products, can be obtained. In addition to conventional MRI for morphological assessment, diffusion-weighted MRI/diffusion tensor imaging is used to evaluate white matter structures in the brain; arterial spin labeling is used for cerebral blood flow evaluation; magnetic resonance elastography for fatty liver and cirrhosis evaluation; magnetic resonance spectroscopy for evaluation of metabolites in specific regions of the brain; and blood oxygenation level-dependent imaging for neurological exploration of eating behavior, obesity, and food perception. This range of applications will continue to expand in the future. Nutritional science is a multidisciplinary and all-inclusive field of research; therefore, there are many different applications of MRI. We present a literature review of MRI techniques that can be used to evaluate the nutritional status, particularly in patients on dialysis. We used MEDLINE as the information source, conducted a keyword search in PubMed, and found that, as a nutritional evaluation method, MRI has been used frequently to comprehensively and quantitatively evaluate muscle mass for the determination of body composition.
Magnetic resonance imaging (MRI) is playing an increasingly important role in evaluating chronic kidney disease (CKD). It has the potential to be used not only for evaluation of physiological and ...pathological states, but also for prediction of disease course. Although different MRI sequences have been employed in renal disease, there are few studies that have compared the different sequences. We compared several multiparametric MRI sequences, and compared their results with the estimated glomerular filtration rate. Principal component analysis showed a similarity between T1 values and tissue perfusion (arterial spin labelling), and between fractional anisotropy (diffusion tensor imaging) and apparent diffusion coefficient values (diffusion-weighted imaging). In multiple regression analysis, only T2* values, derived from the blood oxygenation level-dependent (BOLD) MRI sequence, were associated with estimated glomerular filtration rate slope after adjusting for degree of proteinuria, a classic prognostic factor for CKD. In receiver operating characteristic curve analysis, T2* values were a good predictor of rapid deterioration, regardless of the degree of proteinuria. This suggests further study of the use of BOLD-derived T2* values in the workup of CKD, especially to predict the disease course.
This study performed three-dimensional (3D) magnetic resonance imaging (MRI)-based statistical shape analysis (SSA) by comparing patellofemoral instability (PFI) and normal femur models, and ...developed a machine learning (ML)-based prediction model. Twenty (19 patients) and 31 MRI scans (30 patients) of femurs with PFI and normal femurs, respectively, were used. Bone and cartilage segmentation of the distal femurs was performed and subsequently converted into 3D reconstructed models. The pointwise distance map showed anterior elevation of the trochlea, particularly at the central floor of the proximal trochlea, in the PFI models compared with the normal models. Principal component analysis examined shape variations in the PFI group, and several principal components exhibited shape variations in the trochlear floor and intercondylar width. Multivariate analysis showed that these shape components were significantly correlated with the PFI/non-PFI distinction after adjusting for age and sex. Our ML-based prediction model for PFI achieved a strong predictive performance with an accuracy of 0.909 ± 0.015, and an area under the curve of 0.939 ± 0.009 when using a support vector machine with a linear kernel. This study demonstrated that 3D MRI-based SSA can realistically visualize statistical results on surface models and may facilitate the understanding of complex shape features.
Purpose
We explored the role of histogram analysis of apparent diffusion coefficient (ADC) maps for discriminating uterine carcinosarcoma and endometrial carcinoma.
Materials and Methods
We ...retrospectively evaluated findings in 13 patients with uterine carcinosarcoma and 50 patients with endometrial carcinoma who underwent diffusion‐weighted imaging (b = 0, 500, 1000 s/mm2) at 3T with acquisition of corresponding ADC maps. We derived histogram data from regions of interest drawn on all slices of the ADC maps in which tumor was visualized, excluding areas of necrosis and hemorrhage in the tumor. We used the Mann–Whitney test to evaluate the capacity of histogram parameters (mean ADC value, 5th to 95th percentiles, skewness, kurtosis) to discriminate uterine carcinosarcoma and endometrial carcinoma and analyzed the receiver operating characteristic (ROC) curve to determine the optimum threshold value for each parameter and its corresponding sensitivity and specificity.
Results
Carcinosarcomas demonstrated significantly higher mean vales of ADC, 95th, 90th, 75th, 50th, 25th percentiles and kurtosis than endometrial carcinomas (P < 0.05). ROC curve analysis of the 75th percentile yielded the best area under the ROC curve (AUC; 0.904), sensitivity of 100%, and specificity of 78.0%, with a cutoff value of 1.034 × 10‐3 mm2/s.
Conclusion
Histogram analysis of ADC maps might be helpful for discriminating uterine carcinosarcomas and endometrial carcinomas. J. Magn. Reson. Imaging 2016;43:1301–1307.
Abstract
We evaluated a multiclass classification model to predict estimated glomerular filtration rate (eGFR) groups in chronic kidney disease (CKD) patients using magnetic resonance imaging (MRI) ...texture analysis (TA). We identified 166 CKD patients who underwent MRI comprising Dixon-based T1-weighted in-phase (IP)/opposed-phase (OP)/water-only (WO) images, apparent diffusion coefficient (ADC) maps, and T2* maps. The patients were divided into severe, moderate, and control groups based on eGFR borderlines of 30 and 60 mL/min/1.73 m
2
. After extracting 93 texture features (TFs), dimension reduction was performed using inter-observer reproducibility analysis and sequential feature selection (SFS) algorithm. Models were created using linear discriminant analysis (LDA); support vector machine (SVM) with linear, rbf, and sigmoid kernels; decision tree (DT); and random forest (RF) classifiers, with synthetic minority oversampling technique (SMOTE). Models underwent 100-time repeat nested cross-validation. Overall performances of our classification models were modest, and TA based on T1-weighted IP/OP/WO images provided better performance than those based on ADC and T2* maps. The most favorable result was observed in the T1-weighted WO image using RF classifier and the combination model was derived from all T1-weighted images using SVM classifier with rbf kernel. Among the selected TFs, total energy and energy had weak correlations with eGFR.
Abstract A three-dimensional convolutional neural network model was developed to classify the severity of chronic kidney disease (CKD) using magnetic resonance imaging (MRI) Dixon-based T1-weighted ...in-phase (IP)/opposed-phase (OP)/water-only (WO) imaging. Seventy-three patients with severe renal dysfunction (estimated glomerular filtration rate eGFR < 30 mL/min/1.73 m 2 , CKD stage G4–5); 172 with moderate renal dysfunction (30 ≤ eGFR < 60 mL/min/1.73 m 2 , CKD stage G3a/b); and 76 with mild renal dysfunction (eGFR ≥ 60 mL/min/1.73 m 2 , CKD stage G1–2) participated in this study. The model was applied to the right, left, and both kidneys, as well as to each imaging method (T1-weighted IP/OP/WO images). The best performance was obtained when using bilateral kidneys and IP images, with an accuracy of 0.862 ± 0.036. The overall accuracy was better for the bilateral kidney models than for the unilateral kidney models. Our deep learning approach using kidney MRI can be applied to classify patients with CKD based on the severity of kidney disease.
Endometrial mesonephric-like adenocarcinomas exhibit classical histologic features of mesonephric carcinoma; however, it remains unclear whether these tumors represent mesonephric (Wolffian) ...carcinoma or endometrioid (Müllerian) carcinomas that closely mimic mesonephric carcinoma.
A 32-year-old Japanese primigravida presented with atypical vaginal bleeding. An endometrial biopsy suggested low-grade endometrioid carcinoma, and she was administered medroxyprogesterone acetate. Her tumor recurred 6 years later, and she underwent hysterectomy, salpingo-oophorectomy, and omentectomy, at which point she was diagnosed with mesonephric-like adenocarcinoma of the uterine endometrium. Retrospective pathological review of the initial biopsy confirmed coexisting low-grade endometrioid carcinoma and mesonephric-like adenocarcinoma of the uterine endometrium. On immunohistochemistry, the endometrioid carcinoma component was diffuse positive for estrogen and progesterone receptors but negative for thyroid transcription factor 1. However, the mesonephric-like adenocarcinoma component exhibited a mixture of estrogen receptor- and thyroid transcription factor 1-positive cells within the same glands.
We encountered a patient with coexisting endometrial mesonephric-like adenocarcinoma and low-grade endometrioid carcinoma, which was treated using medroxyprogesterone acetate therapy, resulting in recurrence of mesonephric-like adenocarcinoma alone. These clinicopathological findings support the prevailing notions that mesonephric-like adenocarcinoma is a Müllerian adenocarcinoma exhibiting mesonephric differentiation.
Endometrioid carcinoma is the second most common ovarian tumor, classified as an epithelial-stromal ovarian tumor, and is usually characterized by a cystic tumor with partial solid components on ...magnetic resonance (MR) images. In this case report, we discuss an 81-year-old female who presented with atypical genital bleeding and distended abdomen, for which she underwent abdominal computed tomography and MR imaging. Solid endometrioid carcinoma of the ovary is very rare but was confirmed in our patient during the histological examination after surgery.