Purpose To develop and apply a method to measure in vivo electrical conductivity values using magnetic resonance imaging (MRI) in subjects with breast cancer. Materials and Methods A recently developed technique named MREPT (MR electrical properties tomography) together with a novel coil combination process was used to quantify the conductivity values. The overall technique was validated using a phantom study. In addition, 90 subjects were imaged (50 subjects with previously biopsy-confirmed breast tumor and 40 normal subjects), which was approved by our institutional review board (IRB). A routine clinical protocol, specifically a T₂-weighted FSE (fast spin echo) imaging data, was used for reconstruction of conductivity. Results By employing the coil combination, the relative error in the conductivity map was reduced from ∼70% to 10%. The average conductivity values in breast cancers regions (0.89 ± 0.33S/m) was higher compared to parenchymal tissue (0.43 S/m, P < 0.0001) and fat (0.07 S/m, P < 0.00005) regions. Malignant cases (0.89 S/m, n = 30) showed increased conductivity compared to benign cases (0.56 S/m, n = 5) (P < 0.05). In addition, invasive cancers (0.96 S/m) showed higher mean conductivity compared to in situ cancers (0.57 S/m) (P < 0.0005). Conclusion This study shows that conductivity mapping of breast cancers is feasible using a noninvasive in vivo MREPT technique combined with a coil combination process. The method may provide a tool in the MR diagnosis of breast cancer. J. Magn. Reson. Imaging 2015;42:371–378.

• A new two-stage deep learning approach for efficient microbleeds detection is proposed. • 1st stage detects potential microbleeds candidates, while 2nd stage reduces false positives. • A sensitivity of 93.62% is achieved via 1st stage using high in-plane resolution data. • The average number of false positives per subject is reduced to 1.42 in the 2nd stage. • A validation of using low in-plane resolution data is performed as well.

Magnetic Resonance Imaging (MRI) is sensitive to motion caused by patient movement due to the relatively long data acquisition time. This could cause severe degradation of image quality and therefore affect the overall diagnosis. In this paper, we develop an efficient retrospective 2D deep learning method called stacked U-Nets with self-assisted priors to address the problem of rigid motion artifacts in 3D brain MRI. The proposed work exploits the usage of additional knowledge priors from the corrupted images themselves without the need for additional contrast data. The proposed network learns the missed structural details through sharing auxiliary information from the contiguous slices of the same distorted subject. We further design a refinement stacked U-Nets that facilitates preserving the spatial image details and improves the pixel-to-pixel dependency. To perform network training, simulation of MRI motion artifacts is inevitable. The proposed network is optimized by minimizing the loss of structural similarity (SSIM) using the synthesized motion-corrupted images from 83 real motion-free subjects. We present an intensive analysis using various types of image priors: the proposed self-assisted priors and priors from other image contrast of the same subject. The experimental analysis proves the effectiveness and feasibility of our self-assisted priors since it does not require any further data scans. The overall image quality of the motion-corrected images via the proposed motion correction network significantly improves SSIM from 71.66% to 95.03% and declines the mean square error from 99.25 to 29.76. These results indicate the high similarity of the brain's anatomical structure in the corrected images compared to the motion-free data. The motion-corrected results of both the simulated and real motion data showed the potential of the proposed motion correction network to be feasible and applicable in clinical practices.

Objectives The objective of this study was to develop a computer-aided detection system for automated brain metastases detection using magnetic resonance black-blood imaging and compare its applicability with conventional magnetization-prepared rapid gradient echo (MP-RAGE) imaging. Materials and Methods Twenty-six patients with brain metastases were imaged with a contrast-enhanced, 3-dimensional, whole-brain magnetic resonance black-blood pulse sequence. Approval from the institutional review board and informed consent from the patients were obtained. Preprocessing steps included B1 inhomogeneity correction and brain extraction. The computer-aided detection system used 3-dimensional template matching, which measured normalized cross-correlation coefficient to generate possible metastases candidates. An artificial neural network was used for classification after various volume features were extracted. The same detection procedure was tested with contrast-enhanced MP-RAGE, which was also acquired from the same patients. Results The performance of the proposed detection method was measured by the area under the receiver operating characteristic curve (AUROC), sensitivity, and specificity values. In the black-blood case, detection process displayed an AUROC of 0.9355, a sensitivity value of 81.1%, and a specificity value of 98.2%. Magnetization-prepared rapid gradient echo data showed an AUROC of 0.6508, a sensitivity value of 30.2%, and a specificity value of 99.97%. Conclusions The results demonstrate that accurate automated detection of metastatic brain tumors using contrast-enhanced black-blood imaging sequence is possible compared with using conventional contrast-enhanced MP-RAGE sequence.

Medical image segmentation of tissue abnormalities, key organs, or blood vascular system is of great significance for any computerized diagnostic system. However, automatic segmentation in medical image analysis is a challenging task since it requires sophisticated knowledge of the target organ anatomy. This paper develops an end-to-end deep learning segmentation method called Contextual Multi-Scale Multi-Level Network (CMM-Net). The main idea is to fuse the global contextual features of multiple spatial scales at every contracting convolutional network level in the U-Net. Also, we re-exploit the dilated convolution module that enables an expansion of the receptive field with different rates depending on the size of feature maps throughout the networks. In addition, an augmented testing scheme referred to as Inversion Recovery (IR) which uses logical “OR” and “AND” operators is developed. The proposed segmentation network is evaluated on three medical imaging datasets, namely ISIC 2017 for skin lesions segmentation from dermoscopy images, DRIVE for retinal blood vessels segmentation from fundus images, and BraTS 2018 for brain gliomas segmentation from MR scans. The experimental results showed superior state-of-the-art performance with overall dice similarity coefficients of 85.78%, 80.27%, and 88.96% on the segmentation of skin lesions, retinal blood vessels, and brain tumors, respectively. The proposed CMM-Net is inherently general and could be efficiently applied as a robust tool for various medical image segmentations.