Automatic knee cartilage segmentation from multi-contrast MR images using support vector machine classification with spatial dependencies

Objective. Multi-contrast magnetic resonance (MR) imaging super-resolution (SR) reconstruction is an effective solution for acquiring high-resolution MR images. It utilizes anatomical information from auxiliary contrast images to improve the quality of the target contrast images. However, existing studies have simply explored the relationships between auxiliary contrast and target contrast images but did not fully consider different anatomical information contained in multi-contrast images, resulting in texture details and artifacts unrelated to the target contrast images. Approach. To address these issues, we propose a dual contrast attention-guided multi-frequency fusion (DCAMF) network to reconstruct SR MR images from low-resolution MR images, which adaptively captures relevant anatomical information and processes the texture details and low-frequency information from multi-contrast images in parallel. Specifically, after the feature extraction, a feature selection module based on a dual contrast attention mechanism is proposed to focus on the texture details of the auxiliary contrast images and the low-frequency features of the target contrast images. Then, based on the characteristics of the selected features, a high- and low-frequency fusion decoder is constructed to fuse these features. In addition, a texture-enhancing module is embedded in the high-frequency fusion decoder, to highlight and refine the texture details of the auxiliary contrast and target contrast images. Finally, the high- and low-frequency fusion process is constrained by integrating a deeply-supervised mechanism into the DCAMF network. Main results. The experimental results show that the DCAMF outperforms other state-of-the-art methods. The peak signal-to-noise ratio and structural similarity of DCAMF are 39.02 dB and 0.9771 on the IXI dataset and 37.59 dB and 0.9770 on the BraTS2018 dataset, respectively. The image recovery is further validated in segmentation tasks. Significance. Our proposed SR model can enhance the quality of MR images. The results of the SR study provide a reliable basis for clinical diagnosis and subsequent image-guided treatment.

As two different modalities of medical images, Magnetic Resonance (MR) and Computer Tomography (CT), provide mutually-complementary information to doctors in clinical applications. However, to obtain both images sometimes is cost-consuming and unavailable, particularly for special populations. For example, patients with metal implants are not suitable for MR scanning. Also, it is probably infeasible to acquire multi-contrast MR images during once clinical scanning. In this context, to synthesize needed MR images for patients whose CT images are available becomes valuable. To this end, we present a novel generative network, called CAE-ACGAN, which incorporates the advantages of Variational Auto-Encoder (VAE) and Generative Adversarial Network (GAN) with an auxiliary discriminative classifier network. We apply this network to synthesize multi-contrast MR images from single CT and conduct experiments on brain datasets. Our main contributions can be summarized as follows: 1)We alleviate the problems of images blurriness and mode collapse by integrating the advantages of VAE and GAN; 2) We solve the complicated cross-domain, multi-contrast MR synthesis task using the proposed network; 3) The technique of random-extraction-patches is used to lower the limit of insufficient training data, enabling to obtain promising results even with limited available data; 4) By comparing with other typical networks, we are able to yield nearer-real, higher-quality synthetic MR images, demonstrating the effectiveness and stability of our proposed network.

Knee cartilage and bone segmentation is critical for physicians to analyze and diagnose articular damage and knee osteoarthritis (OA). Deep learning (DL) methods for medical image segmentation have largely outperformed traditional methods, but they often need large amounts of annotated data for model training, which is very costly and time-consuming for medical experts, especially on 3D images. In this paper, we report a new knee cartilage and bone segmentation framework, KCB-Net, for 3D MR images based on sparse annotation. KCB-Net selects a small subset of slices from 3D images for annotation, and seeks to bridge the performance gap between sparse annotation and full annotation. Specifically, it first identifies a subset of the most effective and representative slices with an unsupervised scheme; it then trains an ensemble model using the annotated slices; next, it self-trains the model using 3D images containing pseudo-labels generated by the ensemble method and improved by a bi-directional hierarchical earth mover's distance (bi-HEMD) algorithm; finally, it fine-tunes the segmentation results using the primal-dual Internal Point Method (IPM). Experiments on four 3D MR knee joint datasets (the SKI10 dataset, OAI ZIB dataset, Iowa dataset, and iMorphics dataset) show that our new framework outperforms state-of-the-art methods on full annotation, and yields high quality results for small annotation ratios even as low as 10%.

Magnetic resonance imaging (MRI) tasks often involve multiple contrasts. Recently, numerous deep learning-based multi-contrast MRI super-resolution (SR) and reconstruction methods have been proposed to explore the complementary information from the multi-contrast images. However, these methods either construct parameter-sharing networks or manually design fusion rules, failing to accurately model the correlations between multi-contrast images and lacking certain interpretations. In this paper, we propose a multi-contrast convolutional dictionary (MC-CDic) model under the guidance of the optimization algorithm with a well-designed data fidelity term. Specifically, we bulid an observation model for the multi-contrast MR images to explicitly model the multi-contrast images as common features and unique features. In this way, only the useful information in the reference image can be transferred to the target image, while the inconsistent information will be ignored. We employ the proximal gradient algorithm to optimize the model and unroll the iterative steps into a deep CDic model. Especially, the proximal operators are replaced by learnable ResNet. In addition, multi-scale dictionaries are introduced to further improve the model performance. We test our MC-CDic model on multi-contrast MRI SR and reconstruction tasks. Experimental results demonstrate the superior performance of the proposed MC-CDic model against existing SOTA methods. Code is available at https://github.com/lpcccc-cv/MC-CDic.

Knee osteoarthritis (OA) is characterized by the morphological degeneration of cartilage. Efficient segmentation of cartilage is important for cartilage damage diagnosis and to support therapeutic responses. We present a method for knee cartilage segmentation in magnetic resonance images (MRI). Our method incorporates the Hermite Transform to obtain a hierarchical decomposition of contours which describe knee cartilage shapes. Then, we compute a statistical model of the contour of interest from a set of training images. Thereby, our Hierarchical Active Shape Model (HASM) captures a large range of shape variability even from a small group of training samples, improving segmentation accuracy. The method was trained with a training set of 16- MRI of knee and tested with leave-one-out method.

ObjectiveMild neurocognitive disorders (MND) affect a subset of HIV+ patients under effective combination antiretroviral therapy (cART). In this study, we used an innovative multi-contrast magnetic resonance imaging (MRI) approach at high-field to assess the presence of micro-structural brain alterations in MND+ patients.MethodsWe enrolled 17 MND+ and 19 MND− patients with undetectable HIV-1 RNA and 19 healthy controls (HC). MRI acquisitions at 3T included: MP2RAGE for T1 relaxation times, Magnetization Transfer (MT), T2* and Susceptibility Weighted Imaging (SWI) to probe micro-structural integrity and iron deposition in the brain. Statistical analysis used permutation-based tests and correction for family-wise error rate. Multiple regression analysis was performed between MRI data and (i) neuropsychological results (ii) HIV infection characteristics. A linear discriminant analysis (LDA) based on MRI data was performed between MND+ and MND− patients and cross-validated with a leave-one-out test.ResultsOur data revealed loss of structural integrity and micro-oedema in MND+ compared to HC in the global white and cortical gray matter, as well as in the thalamus and basal ganglia. Multiple regression analysis showed a significant influence of sub-cortical nuclei alterations on the executive index of MND+ patients (p = 0.04 he and R2 = 95.2). The LDA distinguished MND+ and MND− patients with a classification quality of 73% after cross-validation.ConclusionOur study shows micro-structural brain tissue alterations in MND+ patients under effective therapy and suggests that multi-contrast MRI at high field is a powerful approach to discriminate between HIV+ patients on cART with and without mild neurocognitive deficits.

Knee cartilage segmentation has been challenging task for many years. This task is usually connected with two major issues. First object of interest is automatic detection and extraction of knee cartilage shape. Second important issue is detection of osteoarthritis, especially, in early stages. This early deterioration is badly recognizable from native images segmentation significantly contributes to precise localization, detection and extraction of early osteoarthritis. Generally, the knee cartilage automatic segmentation and extraction can be performed by various approaches including edge tracking, intensity-based methods, supervised learning, energy minimization, statistical methods and multiregional segmentation methods. Using of particular segmentation method depends on a compromise which user is willing to accept with respect to robustness, segmentation purpose, computational time, accuracy and level of user interaction. This review is mainly focused on fully automatic segmentation methods bringing the recent informations about modeling of cartilage structure via segmentation approaches.

Osteoarthritis is the most common joint disease without any effective cure to halt its' prevalence. Various medical image segmentation techniques have been proposed to extract knee cartilage from magnetic resonance image in order to identify suitable biomarkers to study the progression of osteoarthritis. In this paper, we propose the use of a random walks knee cartilage segmentation model and analyze the model's accuracy using inter-observer reproducibility. For instance, the proposed model has exhibited promising reproducibility compared to manual segmentation in both normal and diseased categories. In normal cartilage segmentation, the proposed model has shown reproducibility index of 0.93±0.022 while in diseased cartilage segmentation, the proposed model has shown reproducibility index of 0.90±0.049. The results suggest that random walks semi-automated segmentation model reduces the level of ambiguity experienced by manual segmentation model; thus establishing the technique as suitable computerized segmentation technique for knee cartilage segmentation.

Abstract: This work explores a novel approach to accurately identify brain tumors from magnetic resonance imaging (MRI) images by utilizing the complementary capabilities of fuzzy C-Means (FCM) clustering and support vector machines (SVM). Using the soft clustering approach, MRI images are meticulously segregated in the first stage, known as FCM clustering. By giving individual voxels varying degrees of membership, FCM deftly handles the inherent ambiguities and complexity of medical pictures to produce a nuanced representation of tissue kinds. SVM, a potent supervised learning technique for binary classification applications, comes after segmentation. uses this information in the second phase of classification. Through the creation of a hyperplane that divides tumor and non-tumor regions to the greatest extent possible, support vector machines (SVM) are able to optimize the margin between these classes and improve their ability to generalize. We deliberately integrate FCM clustering with SVM classification in order to take use of the complementary characteristics that each of these techniques possesses. While SVM's strong classification skills enable it to fine-tune the identification of tumor regions, FCM's sophisticated segmentation captures the spatial distribution of tissues.

Automatic Segmentation of articular cartilage is important for the quantitative analysis as well as for 2D or 3D visualization of cartilage. Automatic segmentation of cartilage is quite challenging, particularly for OA patients having Bone Marrow Edema (BME) like lesions. Objective of this study was to develop an automatic segmentation approach which can work well for healthy as well as OA patients with and without BME lesions. In this study we proposed an automatic cartilage segmentation approach based upon modification of radial approach and using T2 map values. Accuracy of proposed method was tested with manual segmentation by computing coefficient of similarity. Proposed automated segmentation method was successfully tested on 12 MRI data set. This method provided an accuracy of 89 ± 1 % DSC in patient without BME and 84 ± 1.7 % in subjects having bone lesions. 2D WearMap of both T2 values and thickness values were generated.

A cooperative blind spectrum detection with the Doolittle decomposition and support vector machine (SVM) classification is proposed to improve the performance and reduce the complexity and latency of spectrum detection in hybrid Geostationary Earth Orbit and Low Earth Orbit (GEO-LEO) satellite constellation networks. First, a principal component analysis (PCA) algorithm is adopted to process perceive signals. Second, the maximum-to-minimum eigenvalue ratio of a covariance matrix is utilized to construct statistics with self-learning ability by using the Doolittle decomposition. Third, primary user and noise signals are labeled with “+1” and “$-$1” as training labels, respectively. Finally, a high-performance PCA-SVM classification is generated by training the labels and the corresponding statistics. Innovations mainly include the Doolittle decomposition of the covariance matrix in the SVM and the PCA to preprocess the perceived signals. Simulation results show that the detection probability of the proposed scheme outperforms the energy detection and the SVM algorithm by 55 and 15% at a signal-to-noise ratio (SNR) of $-$15 dB, respectively. The average error probability of the proposed scheme is 20% lower than that of the SVM at an SNR of $-$20 dB. Therefore, in hybrid GEO-LEO satellite constellation networks, the cooperative blind spectrum detection with the Doolittle decomposition and SVM classification can detect the spectrum with low complexity and high performance effectively.

This paper proposes to use border training patterns in order to improve Support Vector Machine (SVM) classification accuracy of hyperspectral images. In the proposed approach, border training patterns which are close to the separating hyperplane, are obtained in two consecutive steps and considered as final training set. In the first step, clustering is performed to the full initial training data of each class. Then, cluster centers of each class are taken as the reduced size training data and forwarded to the second step. In the second step, this reduced size training data is used in the training of SVM and cluster centers which are obtained as support vectors at this step are regarded to be located close to the hyperplane border. Finally, cluster centers which are found as support vectors and original training samples contained in these clusters only are assigned as border training patterns. Experimental results are presented to show that the proposed approach improves SVM classification accuracy.

Segmentation of knee cartilage has been useful for opportune diagnosis and treatment of osteoarthritis (OA). This paper presents a semiautomatic segmentation technique based on Active Shape Models (ASM) combined with Local Binary Patterns (LBP) and its approaches to describe the surrounding texture of femoral cartilage. The proposed technique is tested on a 16-image database of different patients and it is validated through Leave- One-Out method. We compare different segmentation techniques: ASM-LBP, ASM-medianLBP, and ASM proposed by Cootes. The ASM-LBP approaches are tested with different ratios to decide which of them describes the cartilage texture better. The results show that ASM-medianLBP has better performance than ASM-LBP and ASM. Furthermore, we add a routine which improves the robustness versus two principal problems: oversegmentation and initialization.

Nowadays, the Structural Building Health Damage Monitoring System (SBHDMS) is a crucial technology for predicting the civil building structures' health. SBHDMS contains abnormal changes in the buildings in terms of damage levels. Natural Disasters like Earthquakes, Floods, and cyclones affect the unusual changes in the buildings. If the building undergoes any natural disaster, the sensors capture the vibration data or change the buildings' structure. Due to the vibration data, these unusual changes can be analyzed. Here sensors or Machine Learning based Building Damage Prediction (MLBDP) are used for capturing and collecting the vibration data. This paper proposes a Novel Rough Set based Artificial Neural Network with Support Vector Machine (RAS) metaheuristic method. RAS method is used to predict the damaged building's vibration data levels captured by the sensors. For the feature reduction subset, we use one of the essential pre-processing method called the Rough set theory (RST) strategy. RAS has two contributions. The first one is the Support Vector Machine (SVM) classification method used for identifying the structures of the buildings. The artificial Neural Network (ANN) method used to predict the buildings' damage levels is the second contribution. The proposed method (RAS) is accurately predicting the conditions of the construction building structure and predicting the damage levels, without human intervention. Comparing the results states that the proposed method accuracy is better than SVM's classification methods, ANN. The prediction analysis depicts that the RAS method can effectively detect the damage levels.

Detection Traffic Congestion Based on Twitter Data using Machine Learning