Synthetic computed tomography generation for abdominal adaptive radiotherapy using low-field magnetic resonance imaging.

Background and Purpose: Magnetic Resonance guided Radiotherapy (MRgRT) still needs the acquisition of Computed Tomography (CT) images and co-registration between CT and Magnetic Resonance Imaging (MRI). The generation of synthetic CT (sCT) images from the MR data can overcome this limitation. In this study we aim to propose a Deep Learning (DL) based approach for sCT image generation for abdominal Radiotherapy using low field MR images. Materials and methods: CT and MR images were collected from 76 patients treated on abdominal sites. U-Net and conditional Generative Adversarial Network (cGAN) architectures were used to generate sCT images. Additionally, sCT images composed of only six bulk densities were generated with the aim of having a Simplified sCT.Radiotherapy plans calculated using the generated images were compared to the original plan in terms of gamma pass rate and Dose Volume Histogram (DVH) parameters. Results: sCT images were generated in 2 s and 2.5 s with U-Net and cGAN architectures respectively.Gamma pass rates for 2%/2mm and 3%/3mm criteria were 91% and 95% respectively. Dose differences within 1% for DVH parameters on the target volume and organs at risk were obtained. Conclusion: U-Net and cGAN architectures are able to generate abdominal sCT images fast and accurately from low field MRI.

Bladder Wall Segmentation and Characterization on MR Images: Computer-Aided Spina Bifida Diagnosis.

(1) Background: Segmentation of the bladder inner's wall and outer boundaries on Magnetic Resonance Images (MRI) is a crucial step for the diagnosis and the characterization of the bladder state and function. This paper proposes an optimized system for the segmentation and the classification of the bladder wall. (2) Methods: For each image of our data set, the region of interest corresponding to the bladder wall was extracted using LevelSet contour-based segmentation. Several features were computed from the extracted wall on T2 MRI images. After an automatic selection of the sub-vector containing most discriminant features, two supervised learning algorithms were tested using a bio-inspired optimization algorithm. (3) Results: The proposed system based on the improved LevelSet algorithm proved its efficiency in bladder wall segmentation. Experiments also showed that Support Vector Machine (SVM) classifier, optimized by Gray Wolf Optimizer (GWO) and using Radial Basis Function (RBF) kernel outperforms the Random Forest classification algorithm with a set of selected features. (4) Conclusions: A computer-aided optimized system based on segmentation and characterization, of bladder wall on MRI images for classification purposes is proposed. It can significantly be helpful for radiologists as a part of spina bifida study.

Technical note: Design and initial evaluation of a novel physical breast phantom to monitor image quality in digital breast tomosynthesis.

PURPOSE: To describe the creation process of a new breast phantom specifically designed to monitor quality control (QC) metrics consistency over several months in digital breast tomosynthesis (DBT). METHODS: The semi-anthropomorphic Tomomam® phantom was designed and evaluated twice monthly on a single Hologic Selenia Dimensions® unit over 5 months. The phantom is manufactured in a one-piece epoxy resin homogeneous material as the basis for manufacturing, simulating breast tissue as 50% equivalent glandular (GL)/50% equivalent adipose (AD) and compressed thickness of 60 mm. The distribution of test objects on different planes inside the phantom should allow the quantification of 10 image quality metrics: reproducibility, signal difference-to-noise ratio (SDNR), geometric distortions in the plane, missing or added tissue at chest wall, at the top and bottom of images stack and lateral sides, in-plane homogeneity, image scoring, artifact spread function (ASF), geometric distortions in the volume. SDNR was quantified according to GL and AD tissues. Tolerance criteria per parameter were described to analyze results over the study time. RESULTS: Mean scores were equal to 15.4, 15.0, and 11.6 for masses, microcalcifications, and fibers, respectively. A large difference between GL and AD tissues for SDNR metrics was noted over the study time: the best results were obtained from GL tissues. Both geometric distortions and local homogeneity in the plane conformed to expected values. The mean volume value of the triangular prism was 11.3% greater than the expected value due to a reconstruction height equal to 66 mm instead of 60 mm. CONCLUSIONS: In this study, we monitored several QC metrics discriminating GL and AD tissues by using a new breast phantom developed by us. The preliminary clinical tests demonstrated that the Tomomam® phantom could be used to reliably and efficiently track 10 QC metrics with a single acquisition. More data need to be acquired to refine tolerance criteria for some metrics.

Controlling the optical pathlength in continuous-wave reflectance spectroscopy using polarization.

We investigate potential improvements of continuous-wave diffuse reflectance spectroscopy within highly scattering media by employing polarization gating. Simulations are used to show the extent at which the effective optical pathlength varies in a typical scattering medium as a function of the optical wavelength, the total level of absorption, and the selected polarization channels, including elliptical and circular polarization channels. Experiments then demonstrate that a wavelength dependent polarization gating scheme may reduce the prior knowledge required to solve the problem of chromophore quantification. This is achieved by finding combinations of polarization channels which have similar effective optical pathlengths through the medium at each wavelength.

Multi-View Separable Pyramid Network for AD Prediction at MCI Stage by 18F-FDG Brain PET Imaging.

Alzheimer's Disease (AD), one of the main causes of death in elderly people, is characterized by Mild Cognitive Impairment (MCI) at prodromal stage. Nevertheless, only part of MCI subjects could progress to AD. The main objective of this paper is thus to identify those who will develop a dementia of AD type among MCI patients. 18F-FluoroDeoxyGlucose Positron Emission Tomography (18F-FDG PET) serves as a neuroimaging modality for early diagnosis as it can reflect neural activity via measuring glucose uptake at resting-state. In this paper, we design a deep network on 18F-FDG PET modality to address the problem of AD identification at early MCI stage. To this end, a Multi-view Separable Pyramid Network (MiSePyNet) is proposed, in which representations are learned from axial, coronal and sagittal views of PET scans so as to offer complementary information and then combined to make a decision jointly. Different from the widely and naturally used 3D convolution operations for 3D images, the proposed architecture is deployed with separable convolution from slice-wise to spatial-wise successively, which can retain the spatial information and reduce training parameters compared to 2D and 3D networks, respectively. Experiments on ADNI dataset show that the proposed method can yield better performance than both traditional and deep learning-based algorithms for predicting the progression of Mild Cognitive Impairment, with a classification accuracy of 83.05%.

Multiscale spatial gradient features for 18F-FDG PET image-guided diagnosis of Alzheimer's disease.

BACKGROUND AND OBJECTIVE: 18F-FluoroDeoxyGlucose Positron Emission Tomography (18F-FDG PET) is one of the imaging biomarkers to diagnose Alzheimer's Disease (AD). In 18F-FDG PET images, the changes of voxels' intensities reflect the differences of glucose rates, therefore voxel intensity is usually used as a feature to distinguish AD from Normal Control (NC), or at earlier stage to distinguish between progressive and stable Mild Cognitive Impairment (pMCI and sMCI). In this paper, 18F-FDG PET images are characterized in an alternative way-the spatial gradient, which is motivated by the observation that the changes of 18F-FDG rates also cause gradient changes. METHODS: We improve Histogram of Oriented Gradient (HOG) descriptor to quantify spatial gradients, thereby achieving the goal of diagnosing AD. First, the spatial gradient of 18F-FDG PET image is computed, and then each subject is segmented into different regions by using an anatomical atlas. Second, two types of improved HOG features are extracted from each region, namely Small Scale HOG and Large Scale HOG, then some relevant regions are selected based on a classifier fed with spatial gradient features. Last, an ensemble classification framework is designed to make a decision, which considers the performance of both individual and concatenated selected regions. RESULTS: the evaluation is done on ADNI dataset. The proposed method outperforms other state-of-the-art 18F-FDG PET-based algorithms for AD vs. NC with an accuracy, a sensitivity and a specificity values of 93.65%, 91.22% and 96.25%, respectively. For the case of pMCI vs. sMCI, the three metrics are 75.38%, 74.84% and 77.11%, which is significantly better than most existing methods. Besides, promising results are also achieved for multiple classifications under 18F-FDG PET modality. CONCLUSIONS: 18F-FDG PET images can be characterized by spatial gradient features for diagnosing AD and its early stage, and the proposed ensemble framework can enhance the classification performance.

Influence of normalization and color features on super-pixel classification: application to cytological image segmentation.

Super-pixel feature extraction is a key problem to get an acceptable performance in color super-pixel classification. Given a color feature extraction problem, it is necessary to know which is the best approach to solve this problem. In the current work, we're interested in the challenge of nucleus and cytoplasm automatic recognition in the cytological image. We propose an automatic process for white blood cells (WBC) segmentation using super-pixel classification. The process is divided into five steps. In first step, the color normalization is calculated. The super-pixels generation by Simple Linear Iterative Clustering algorithm is performed in the second step. In third step, the color property is used to achieve illumination invariance. In fourth step, color features are calculated on each super-pixel. Finally, supervised learning is realized to classify each super-pixel into nucleus and cytoplasm region. The present work rallied an exhaustive statistical evaluation of a very wide variety of the color super-pixel classification, with height normalization methods, four-color spaces and four feature extraction techniques. Normalization and color spaces slightly increase the average accuracy of super-pixel classification. Our experiments based to statistical comparison allow to conclude that comprehensive gray world normalized normalization is better than without normalization for super-pixel classification achieving the first positions in the Friedman ranking. RGB space is the best color spaces to be used in super-pixel feature extraction for nucleus and cytoplasm segmentation. For feature extraction, the learning methods work better on the first order statistics features for the automatic WBC segmentation.

Multilevel Feature Representation of FDG-PET Brain Images for Diagnosing Alzheimer's Disease.

Using a single imaging modality to diagnose Alzheimer's disease (AD) or mild cognitive impairment (MCI) is a challenging task. FluoroDeoxyGlucose Positron Emission Tomography (FDG-PET) is an important and effective modality used for that purpose. In this paper, we develop a novel method by using single modality (FDG-PET) but multilevel feature, which considers both region properties and connectivities between regions to classify AD or MCI from normal control. First, three levels of features are extracted: statistical, connectivity, and graph-based features. Then, the connectivity features are decomposed into three different sets of features according to a proposed similarity-driven ranking method, which can not only reduce the feature dimension but also increase the classifier's diversity. Last, after feeding the three levels of features to different classifiers, a new classifier selection strategy, maximum Mean squared Error (mMsE), is developed to select a pair of classifiers with high diversity. In order to do the majority voting, a decision-making scheme, a nested cross validation technique is applied to choose another classifier according to the accuracy. Experiments on Alzheimer's Disease Neuroimaging Initiative database show that the proposed method outperforms most FDG-PET-based classification algorithms, especially for classifying progressive MCI (pMCI) from stable MCI (sMCI).

Histogram-Based Features Selection and Volume of Interest Ranking for Brain PET Image Classification.

Positron emission tomography (PET) is a molecular medical imaging modality which is commonly used for neurodegenerative diseases diagnosis. Computer-aided diagnosis, based on medical image analysis, could help quantitative evaluation of brain diseases such as Alzheimer's disease (AD). A novel method of ranking the effectiveness of brain volume of interest (VOI) to separate healthy control from AD brains PET images is presented in this paper. Brain images are first mapped into anatomical VOIs using an atlas. Histogram-based features are then extracted and used to select and rank VOIs according to the area under curve (AUC) parameter, which produces a hierarchy of the ability of VOIs to separate between groups of subjects. The top-ranked VOIs are then input into a support vector machine classifier. The developed method is evaluated on a local database image and compared to the known selection feature methods. Results show that using AUC outperforms classification results in the case of a two group separation.

Bayesian method with spatial constraint for retinal vessel segmentation.

A Bayesian method with spatial constraint is proposed for vessel segmentation in retinal images. The proposed model makes the assumption that the posterior probability of each pixel is dependent on posterior probabilities of their neighboring pixels. An energy function is defined for the proposed model. By applying the modified level set approach to minimize the proposed energy function, we can identify blood vessels in the retinal image. Evaluation of the developed method is done on real retinal images which are from the DRIVE database and the STARE database. The performance is analyzed and compared to other published methods using a number of measures which include accuracy, sensitivity, and specificity. The proposed approach is proved to be effective on these two databases. The average accuracy, sensitivity, and specificity on the DRIVE database are 0.9529, 0.7513, and 0.9792, respectively, and for the STARE database 0.9476, 0.7147, and 0.9735, respectively. The performance is better than that of other vessel segmentation methods.

Automatic segmentation and measurement of vasculature in retinal fundus images using probabilistic formulation.

The automatic analysis of retinal blood vessels plays an important role in the computer-aided diagnosis. In this paper, we introduce a probabilistic tracking-based method for automatic vessel segmentation in retinal images. We take into account vessel edge detection on the whole retinal image and handle different vessel structures. During the tracking process, a Bayesian method with maximum a posteriori (MAP) as criterion is used to detect vessel edge points. Experimental evaluations of the tracking algorithm are performed on real retinal images from three publicly available databases: STARE (Hoover et al., 2000), DRIVE (Staal et al., 2004), and REVIEW (Al-Diri et al., 2008 and 2009). We got high accuracy in vessel segmentation, width measurements, and vessel structure identification. The sensitivity and specificity on STARE are 0.7248 and 0.9666, respectively. On DRIVE, the sensitivity is 0.6522 and the specificity is up to 0.9710.

Semi-automated detection of the foveal avascular zone in fluorescein angiograms in diabetes mellitus.

BACKGROUND: The foveal avascular zone (FAZ) is known to enlarge in diabetic retinopathy. In a preliminary study, the authors applied a region growing algorithm to fluorescein angiograms to detect the FAZ in a semi-automated fashion. METHODS: The FAZ in 44 fluorescein angiograms of 44 eyes of 41 patients with diabetic retinopathy underwent manual outlining, then analysis with the region growing function of the ENVI image analysis software. The same algorithm was applied after median filtering of the images. RESULTS: Correlation coefficient was 0.98 between the first two authors, 0.89 between the first author and semi-automated detection before median filtering and 0.91 after median filtering. Average surface areas however, were smaller with semi-automated detection (1951 pixels before and 2226 pixels after median filtering) than with manual detection (3032 pixels). CONCLUSIONS: Semi-automated detection of the FAZ is possible, but refinements need to be made in angiogram quality and/or image pretreatment to improve results.