Texture classification using refined histogram

Detection and Spatiotemporal Analysis of In-vitro 3D Migratory Triple-Negative Breast Cancer Cells

The invasion of cancer cells into the surrounding tissues is one of the hallmarks of cancer. However, a precise quantitative understanding of the spatiotemporal patterns of cancer cell migration and invasion still remains elusive. A promising approach to investigate these patterns are 3D cell cultures, which provide more realistic models of cancer growth compared to conventional 2D monolayers. Quantifying the spatial distribution of cells in these 3D cultures yields great promise for understanding the spatiotemporal progression of cancer. In the present study, we present an image processing and segmentation pipeline for the detection of 3D GFP-fluorescent triple-negative breast cancer cell nuclei, and we perform quantitative analysis of the formed spatial patterns and their temporal evolution. The performance of the proposed pipeline was evaluated using experimental 3D cell culture data, and was found to be comparable to manual segmentation, outperforming four alternative automated methods. The spatiotemporal statistical analysis of the detected distributions of nuclei revealed transient, non-random spatial distributions that consisted of clustered patterns across a wide range of neighbourhood distances, as well as dispersion for larger distances. Overall, the implementation of the proposed framework revealed the spatial organization of cellular nuclei with improved accuracy, providing insights into the 3 dimensional inter-cellular organization and its progression through time.

Laser stripe segmentation and centerline extraction based on 3D scanning imaging

Ambient noise and illumination inhomogeneity will seriously affect the high-precision measurement of structured light 3D morphology. To overcome the influences of these factors, a new, to the best of our knowledge, sub-pixel extraction method for the center of laser stripes is proposed. First, an automatic segmentation model of structured light stripe based on the UNet deep learning network and level set is constructed. Coarse segmentation of laser stripes using the UNet network can effectively segment more complex scenes and automatically obtain a prior shape information. Then, the prior information is used as a shape constraint for fine segmentation of the level set, and the energy function of the level set is improved. Finally, the stripe normal field is obtained by calculating the stripe gradient vector, and the center of the stripe is extracted by fusing the gray center of gravity method according to the normal direction of the stripe distribution. The experimental results show that the average width error of different rows of point cloud data of workpieces with different widths is less than 0.3 mm, and the average repeatability extraction error is less than 0.2 mm.

The association of body composition with abdominal aortic aneurysm growth after endovascular aneurysm repair

Background

Body composition (BC) may be associated with abdominal aortic aneurysm (AAA) growth, but the results of previous research are contradictory. This study aimed to explore the relationship between BC and postoperative aneurysm progression.

Methods

Patients with regular postoperative follow-ups were retrospectively identified. The volume change of the aneurysm was measured to evaluate AAA progression. After segmenting different body components (subcutaneous fat, visceral fat, pure muscle, and intramuscular fat), the shape features and gray features of these tissues were extracted. Uni- and multivariable methods were used to analyze the relationship between imaging features of BC and AAA growth.

Results

A total of 94 patients (68 ± 8 years) were eligible for feature analyses. Patients with expansive aneurysms (29/94; volume change > 2%) were classified into Group(+) and others with stable or shrunken aneurysms (65/94) were classified into Group(−). Compared with Group(+), Group(−) showed a higher volume percent of pure muscle (21.85% vs 19.51%; p = .042) and a lower value of intramuscular fat (1.23% vs 1.65%; p = .025). CT attenuation of muscle tissues of Group(−) got a higher mean value (31.16 HU vs 23.92 HU; p = .019) and a lower standard deviation (36.12 vs 38.82; p = .006) than Group(+). For adipose tissue, we found no evidence of a difference between the two groups. The logistic regression model containing muscle imaging features showed better discriminative accuracy than traditional factors (84% vs 73%).

Conclusions

Muscle imaging features are associated with the volume change of postoperative aneurysms and can make an early prediction. Adipose tissue is not specifically related to AAA growth.

Quantitative Evaluation of Retinal Microvascular Abnormalities in Patients With Type 2 Diabetes Mellitus Without Clinical Sign of Diabetic Retinopathy

Purpose

To evaluate microvascular abnormalities in the macula and peripapillary area in diabetic patients without clinical signs of diabetic retinopathy (DR) and compare them with healthy control eyes, using optical coherence tomography angiography (OCTA).

Methods

A prospective study was performed of 49 eyes from 49 diabetic patients without clinical signs of DR and a control group of 52 eyes from 52 healthy normal individuals. The 3 × 3 mm macular scans and 4.5 × 4.5 mm optic disc scans were obtained with the OCTA RTVue-XR Avanti system. Angiograms from the superficial capillary plexus, the deep capillary plexus of the macula scans, and radial peripapillary capillary plexus of the optic disc scans were analyzed with MATLAB. Multivariate binary logistic regression and the least absolute shrinkage and selection operator (LASSO) regression were used to select ideal parameters that distinguish diabetic eyes without DR from normal eyes. A receiver operating characteristic (ROC) curve was generated, and sensitivity and specificity were calculated.

Results

Our final model identified FD-300 (foveal vessel density in a 300-µm-wide region around foveal avascular zone) as the only parameter selected by both the LASSO regression and the final multivariate logistic regression model that significantly differentiates diabetic eyes without clinical signs of DR from healthy normal eyes. The area under the ROC curve of FD-300 was 0.685, and sensitivity and specificity were 65.3% and 71.2%, respectively.

Conclusions

Quantitative evaluation of retinal microvascular abnormalities using OCTA identified FD-300 as a useful biomarker versus the other macular and peripapillary OCTA metrics in the early detection of preclinical diabetic retinal abnormalities.

Translational Relevance

OCTA may be useful in detecting early retinal microvascular abnormalities in diabetic patients before the clinical findings of DR become visible.

Detection of the Microvascular Changes of Diabetic Retinopathy Progression Using Optical Coherence Tomography Angiography

Purpose

To investigate microvascular parameters that are related to the severity of diabetic retinopathy (DR) with optical coherence tomography angiography (OCTA).

Methods

In total, 105 eyes from 105 diabetic patients were recruited in this prospective cross-sectional study, including 37 eyes with no clinical signs of DR (NoDR), 43 eyes with nonproliferative diabetic retinopathy (NPDR), and 25 eyes with proliferative diabetic retinopathy (PDR). Angiogram images from the parafoveal superficial capillary plexus (SCP), the deep capillary plexus (DCP), and the radial peripapillary capillary plexus were analyzed, and metrics were compared among groups. Multivariate regression analysis was used to identify the best OCTA parameters that could distinguish DR severity among groups.

Results

Parafoveal vessel diameter index in the SCP and vessel density (VD) in the DCP showed the strongest correlation with the severity of DR (P < 0.01). Extrafoveal avascular area in the SCP was the parameter that could most distinguish NoDR from NPDR (P < 0.01) with sensitivity and specificity of 83.72% and 78.38%, respectively. VD in the DCP also was the most sensitive biomarker to distinguish NPDR from PDR (P < 0.01) with sensitivity and specificity of 84.00% and 79.07%, respectively.

Conclusions

The microvascular changes in the SCP and DCP in DR may have different characteristics that could be identified with specific OCTA parameters. OCTA serves as a promising technology to discriminate eyes with different severity of DR.

Translational Relevance

Our study investigated OCTA metrics and severity of DR. At different stages of DR, ophthalmologists may focus on specific OCTA parameters to predict the progression of retinopathy in individual patients.

Transient Support from Fibroblasts is Sufficient to Drive Functional Vascularization in Engineered Tissues

Formation of capillary blood vasculature is a critical requirement for native as well as engineered organs and can be induced in vitro by co-culturing endothelial cells with fibroblasts. However, whether these fibroblasts are required only in the initial morphogenesis of endothelial cells or needed throughout is unknown, and the ability to remove these stromal cells after assembly could be useful for clinical translation. In this study, we introduce a technique termed CAMEO (Controlled Apoptosis in Multicellular Tissues for Engineered Organogenesis), whereby fibroblasts are selectively ablated on demand, and utilize it to probe the dispensability of fibroblasts in vascular morphogenesis. The presence of fibroblasts is shown to be necessary only during the first few days of endothelial cell morphogenesis, after which they can be ablated without significantly affecting the structural and functional features of the developed vasculature. Furthermore, we demonstrate the use of CAMEO to vascularize a construct containing primary human hepatocytes that improved tissue function. In conclusion, this study suggests that transient, initial support from fibroblasts is sufficient to drive vascular morphogenesis in engineered tissues, and this strategy of engineering-via-elimination may provide a new general approach for achieving desired functions and cell compositions in engineered organs.

Glass-cutting medical images via a mechanical image segmentation method based on crack propagation

Medical image segmentation is crucial in diagnosing and treating diseases, but automatic segmentation of complex images is very challenging. Here we present a method, called the crack propagation method (CPM), based on the principles of fracture mechanics. This unique method converts the image segmentation problem into a mechanical one, extracting the boundary information of the target area by tracing the crack propagation on a thin plate with grooves corresponding to the area edge. The greatest advantage of CPM is in segmenting images involving blurred or even discontinuous boundaries, a task difficult to achieve by existing auto-segmentation methods. The segmentation results for synthesized images and real medical images show that CPM has high accuracy in segmenting complex boundaries. With increasing demand for medical imaging in clinical practice and research, this method will show its unique potential.

Automatic segmentation of complex medical images is challenging. Here, the authors present a crack propagation method based on the principles of fracture mechanics: extracting the boundary information of the target area by tracing the crack propagation on a thin plate with grooves corresponding to the area edge.

Semi-implicit level set formulation for lithographic source and mask optimization

The convergence of lithographic source and mask optimization (SMO) has been plagued by the prohibitive time-step dictated by the stability of the explicit Euler-forward scheme in the gradient-based optimization procedure. As a remedy, we solve the distance level-set regularized reformulation of the SMO by discretizing the stability-relevant terms in an implicit manner and apply operator splitting to separately update source and mask patterns in coordinate dimensions by solving the tridiagonal systems of linear equations using the Thomas method, combining stability and simplicity. Simulation results merit the superiority of the proposed SMO approach with improved convergence by overcoming the stability constraints of the Courant-Friedrichs-Lewy (CFL) condition.

Multi-phase level set algorithm based on fully convolutional networks (FCN-MLS) for retinal layer segmentation in SD-OCT images with central serous chorioretinopathy (CSC)

As a function of the spatial position of the optical coherence tomography (OCT) image, retinal layer thickness is an important diagnostic indicator for many retinal diseases. Reliable segmentation of the retinal layer is necessary for extracting useful clinical information. However, manual segmentation of these layers is time-consuming and prone to bias. Furthermore, due to speckle noise, low image contrast, retinal detachment, and also irregular morphological features make the automatic segmentation task challenging. To alleviate these challenges, in this paper, we propose a new coarse-fine framework combining the full convolutional network (FCN) with a multiphase level set (named FCN-MLS) for automatic segmentation of nine boundaries in retinal spectral OCT images. In the coarse stage, FCN is used to learn the characteristics of specific retinal layer boundaries and achieve classification of four retinal layers. The boundaries are then extracted and the remaining boundaries are initialized based on a priori information about the thickness of the retinal layer. In order to prevent the overlapping of the segmentation interfaces, a regional restriction technique is used in the multi-phase level to evolve the boundaries to achieve fine nine retinal layers segmentation. Experimental results on 1280 B-scans show that the proposed method can segment nine retinal boundaries accurately. Compared with the manual delineation, the overall mean absolute boundary location difference and the overall mean absolute thickness difference were 5.88 ± 2.38μm and 5.81 ± 2.19μm, which showed a good consistency with manual segmentation by the physicians. Our experimental results also outperform state-of-the-art methods.

Automated drosophila heartbeat counting based on image segmentation technique on optical coherence tomography

Drosophila and human cardiac genes are very similar. Biological parametric studies on drosophila cardiac have improved our understanding of human cardiovascular disease. Drosophila cardiac consist of five circular chambers: a conical chamber (CC) and four ostia sections (O1-O4). Due to noise and grayscale discontinuity on optical coherence tomography (OCT) images, previous researches used manual counting or M-mode to analyze heartbeats, which are inefficient and time-consuming. An automated drosophila heartbeat counting algorithm based on the chamber segmentation is developed for OCT in this study. This algorithm has two parts: automated chamber segmentation and heartbeat counting. In addition, this study proposes a principal components analysis (PCA)-based supervised learning method for training the chamber contours to make chamber segmentation more accurate. The mean distances between the conical, second and third chambers attained by the proposed algorithm and the corresponding manually delineated boundaries defined by two experts were 1.26 ± 0.25, 1.47 ± 1.25 and 0.84 ± 0.60 (pixels), respectively. The area overlap similarities were 0.83 ± 0.09, 0.75 ± 0.11 and 0.74 ± 0.12 (pixels), respectively. The average calculated heart rates of two-week and six-week drosophila were about 4.77 beats/s and 4.73 beats/s, respectively, which was consistent with the results of manual counting.

Optically guided level set for underwater object segmentation

Combining underwater optical imaging principles and the level set, this paper proposes a novel type of level set method called optically guided level set. This novel method can transform optical challenges in underwater environments (such as the illumination bias and wavelength-selective absorption) into valuable guidance for underwater object segmentation. Using the underwater optical guidance, our novel method can generate accurate object segmentation results by suitable initialization and regular evolving of the level set. The optical guidance core lies in two observations pertaining to the underwater optical imaging process: (i) the overlap between the object region and optical collimation region and (ii) the correspondence between the object structure and irradiation distribution inside the optical collimation. The high accuracy of our proposed method is demonstrated via comparisons to the state-of-the-art level set and salient object detection methods for public underwater images collected in diverse environments. Moreover, by using the work presented in this paper, we plan to demonstrate optical principles' potential for improving computer vision research, which is a promising research topic with many practical applications.

Multiview Contouring for Breast Tumor on Magnetic Resonance Imaging

The shape and contour of the lesion are shown to be effective features for physicians to identify breast tumor as benign or malignant. The region of the lesion is usually manually created by the physician according to their clinical experience; therefore, contouring tumors on breast magnetic resonance imaging (MRI) is difficult and time-consuming. For this purpose, an automatic contouring method for breast tumors was developed for less burden in the analysis and to decrease the observed bias to help in making decisions clinically. In this study, a multiview segmentation method for detecting and contouring breast tumors in MRI was represented. The preprocessing of the proposed method reduces any amount of noises but preserves the shape and contrast of the breast tumor. The two-dimensional (2D) level-set segmentation method extracts contours of breast tumors from the transverse, coronal, and sagittal planes. The obtained contours are further utilized to generate appropriate three-dimensional (3D) contours. Twenty breast tumor cases were evaluated and the simulation results show that the proposed contouring method was an efficient method for delineating 3D contours of breast tumors in MRI.

Segmentation of Heavily Clustered Nuclei from Histopathological Images

Automated cell nucleus segmentation is the key to gain further insight into cell features and functionality which support computer-aided pathology in early diagnosis of diseases such as breast cancer and brain tumour. Despite considerable advances in automated segmentation, it still remains a challenging task to split heavily clustered nuclei due to intensity variations caused by noise and uneven absorption of stains. To address this problem, we propose a novel method applicable to variety of histopathological images stained for different proteins, with high speed, accuracy and level of automation. Our algorithm is initiated by applying a new locally adaptive thresholding method on watershed regions. Followed by a new splitting technique based on multilevel thresholding and the watershed algorithm to separate clustered nuclei. Finalized by a model-based merging step to eliminate oversegmentation and a model-based correction step to improve segmentation results and eliminate small objects. We have applied our method to three image datasets: breast cancer stained for hematoxylin and eosin (H&E), Drosophila Kc167 cells stained for DNA to label nuclei, and mature neurons stained for NeuN. Evaluated results show our method outperforms the state-of-the-art methods in terms of accuracy, precision, F1-measure, and computational time.

Automated segmentation of trabecular and cortical bone from proton density weighted MRI of the knee

Patient-specific implant design and pre- and intra-operative planning is becoming increasingly important in the orthopaedic field. For clinical feasibility of these techniques, fast and accurate segmentation of bone structures from MRI is essential. However, manual segmentation is time intensive and subject to inter- and intra-observer variation. The challenge in developing automatic segmentation algorithms for MRI data mainly exists in the inhomogeneity problem and the low contrast among cortical bone and adjacent tissues. In this paper, we proposed a method for automatic segmentation of knee bone structures for MRI data with a 3D local intensity clustering-based level set and a novel approach to determine the cortical boundary utilizing the normal vector of the trabecular surface. Application to clinical imaging data shows that our method is robust to MRI inhomogeneity. In comparing our method to manual segmentation in 18 femurs and tibiae, we found a dice similarity coefficient (DSC) of 0.9611 ± 0.0052 for the femurs and 0.9591 ± 0.0173 for tibiae. The average surface distance error was 0.4649 ± 0.1430 mm for the femurs and 0.4712 ± 0.2113 mm for the tibiae. The results of the automatic technique thus strongly corresponded to the manual segmentation using less than 3% of the time and with virtually no workload.

Graphical abstract

Computational delineation and quantitative heterogeneity analysis of lung tumor on 18F-FDG PET for radiation dose-escalation

Quantitative measurement and analysis of tumor metabolic activities could provide a more optimal solution to personalized accurate dose painting. We collected PET images of 58 lung cancer patients, in which the tumor exhibits heterogeneous FDG uptake. We design an automated delineation and quantitative heterogeneity measurement of the lung tumor for dose-escalation. For tumor delineation, our algorithm firstly separates the tumor from its adjacent high-uptake tissues using 3D projection masks; then the tumor boundary is delineated with our stopping criterion of joint gradient and intensity affinities. For dose-escalation, tumor sub-volumes with low, moderate and high metabolic activities are extracted and measured. Based on our quantitative heterogeneity measurement, a sub-volume oriented dose-escalation plan is implemented in intensity modulated radiation therapy (IMRT) planning system. With respect to manual tumor delineations by two radiation oncologists, the paired t-test demonstrated our model outperformed the other computational methods in comparison (p < 0.05) and reduced the variability between inter-observers. Compared to standard uniform dose prescription, the dosimetry results demonstrated that the dose-escalation plan statistically boosted the dose delivered to high metabolic tumor sub-volumes (p < 0.05). Meanwhile, the doses received by organs-at-risk (OAR) including the heart, ipsilateral lung and contralateral lung were not statistically different (p > 0.05).

Lithographic source and mask optimization with narrow-band level-set method

Source and mask optimization (SMO) remains a key technique to improve the wafer image printability for technology nodes of 22 nm and beyond, enabling the continuation of the immersion lithography. In this paper, we propose a distance level-set regularized reformulation of the SMO maintaining the desired signed distance property, which secures stable curve evolution and accurate computation with a simpler and more efficient numerical implementation. Consequently, computation load caused by convolution operations and memory requirements of the electric-field caching technique (EFCT) is significantly eased by performing computation only in the narrow band; moreover, the convergence of the updating process is further improved by applying larger Euler time steps of the Courant-Friedrichs-Lewy (CFL) condition with reduced optimization dimensionality. Simulation results of the proposed narrow-band level-set based SMO prove to improve the computation efficiency, memory usage and imaging performance of the full domain methods.

Trajectory energy minimization for cell growth tracking and genealogy analysis

Cell growth experiments with a microfluidic device produce large-scale time-lapse image data, which contain important information on cell growth and patterns in their genealogy. To extract such information, we propose a scheme to segment and track bacterial cells automatically. In contrast with most published approaches, which often split segmentation and tracking into two independent procedures, we focus on designing an algorithm that describes cell properties evolving between consecutive frames by feeding segmentation and tracking results from one frame to the next one. The cell boundaries are extracted by minimizing the distance regularized level set evolution (DRLSE) model. Each individual cell was identified and tracked by identifying cell septum and membrane as well as developing a trajectory energy minimization function along time-lapse series. Experiments show that by applying this scheme, cell growth and division can be measured automatically. The results show the efficiency of the approach when testing on different datasets while comparing with other existing algorithms. The proposed approach demonstrates great potential for large-scale bacterial cell growth analysis.

A Computational Model for Texture Analysis in Images with Fractional Differential Filter for Texture Detection

This paper is dedicated to the modelling of textured images influenced by fractional derivatives for texture detection. As most of the images contain textures, texture analysis becomes the most important for image understanding and it is a key solution for many computer vision applications. Hence, texture must be suitably detected and represented. Nevertheless, existing texture detection algorithms consider texture as a unique feature from edges. The proposed model explores a novel way of developing texture detection algorithm by mimicking edge detection algorithms. The method assumes that texture feature is analogous to edges and thus, the time complexity is reduced significantly. The model proposed in this work is based on Gaussian kernel smoothing, Fractional partial derivatives and a statistical approach. It is justified to be robust to noisy images and possesses statistical interpretation. The model is validated by the classification experiments on different types of textured images from Brodatz album. It achieves higher classification accuracy than the existing methods.

Automated Robust Image Segmentation: Level Set Method Using Nonnegative Matrix Factorization with Application to Brain MRI

We address the problem of fully automated region discovery and robust image segmentation by devising a new deformable model based on the level set method (LSM) and the probabilistic nonnegative matrix factorization (NMF). We describe the use of NMF to calculate the number of distinct regions in the image and to derive the local distribution of the regions, which is incorporated into the energy functional of the LSM. The results demonstrate that our NMF-LSM method is superior to other approaches when applied to synthetic binary and gray-scale images and to clinical magnetic resonance images (MRI) of the human brain with and without a malignant brain tumor, glioblastoma multiforme. In particular, the NMF-LSM method is fully automated, highly accurate, less sensitive to the initial selection of the contour(s) or initial conditions, more robust to noise and model parameters, and able to detect as small distinct regions as desired. These advantages stem from the fact that the proposed method relies on histogram information instead of intensity values and does not introduce nuisance model parameters. These properties provide a general approach for automated robust region discovery and segmentation in heterogeneous images. Compared with the retrospective radiological diagnoses of two patients with non-enhancing grade 2 and 3 oligodendroglioma, the NMF-LSM detects earlier progression times and appears suitable for monitoring tumor response. The NMF-LSM method fills an important need of automated segmentation of clinical MRI.

Computer-Aided Diagnosis with Deep Learning Architecture: Applications to Breast Lesions in US Images and Pulmonary Nodules in CT Scans

This paper performs a comprehensive study on the deep-learning-based computer-aided diagnosis (CADx) for the differential diagnosis of benign and malignant nodules/lesions by avoiding the potential errors caused by inaccurate image processing results (e.g., boundary segmentation), as well as the classification bias resulting from a less robust feature set, as involved in most conventional CADx algorithms. Specifically, the stacked denoising auto-encoder (SDAE) is exploited on the two CADx applications for the differentiation of breast ultrasound lesions and lung CT nodules. The SDAE architecture is well equipped with the automatic feature exploration mechanism and noise tolerance advantage, and hence may be suitable to deal with the intrinsically noisy property of medical image data from various imaging modalities. To show the outperformance of SDAE-based CADx over the conventional scheme, two latest conventional CADx algorithms are implemented for comparison. 10 times of 10-fold cross-validations are conducted to illustrate the efficacy of the SDAE-based CADx algorithm. The experimental results show the significant performance boost by the SDAE-based CADx algorithm over the two conventional methods, suggesting that deep learning techniques can potentially change the design paradigm of the CADx systems without the need of explicit design and selection of problem-oriented features.

Feature-based active contour model and occluding object detection

This paper presents a method for image segmentation and object detection. The proposed strategy consists of two major stages. The first one corresponds to image segmentation, which is based on the active contour model (ACM) algorithm, using an automatic selection of the best candidate features among gradient, polarity, and depth, coupled with a combination of them by the kernel support vector machine (KSVM). Although existing techniques, such as the ones based on ACM, perform well in the single-object case and non-noisy environments, these techniques fail when the scene consists of multiple occluding objects, with possibly similar colors. Thus, the second stage corresponds to the identification of salient and occluded objects based on the fuzzy C-mean algorithm (FCM). In this stage, the depth is included as another clue that allows us to estimate the cluster number and to make the clustering process more robust. In particular, complex occlusions can be handled this way, and the objects can be properly segmented and identified. Experimental results on real images and on several standard datasets have shown the success and effectiveness of the proposed method.

Tracking tumor boundary in MV-EPID images without implanted markers: A feasibility study

PURPOSE

To develop a markerless tracking algorithm to track the tumor boundary in megavoltage (MV)-electronic portal imaging device (EPID) images for image-guided radiation therapy.

METHODS

A level set method (LSM)-based algorithm is developed to track tumor boundary in EPID image sequences. Given an EPID image sequence, an initial curve is manually specified in the first frame. Driven by a region-scalable energy fitting function, the initial curve automatically evolves toward the tumor boundary and stops on the desired boundary while the energy function reaches its minimum. For the subsequent frames, the tracking algorithm updates the initial curve by using the tracking result in the previous frame and reuses the LSM to detect the tumor boundary in the subsequent frame so that the tracking processing can be continued without user intervention. The tracking algorithm is tested on three image datasets, including a 4-D phantom EPID image sequence, four digitally deformable phantom image sequences with different noise levels, and four clinical EPID image sequences acquired in lung cancer treatment. The tracking accuracy is evaluated based on two metrics: centroid localization error (CLE) and volume overlap index (VOI) between the tracking result and the ground truth.

RESULTS

For the 4-D phantom image sequence, the CLE is 0.23 ± 0.20 mm, and VOI is 95.6% ± 0.2%. For the digital phantom image sequences, the total CLE and VOI are 0.11 ± 0.08 mm and 96.7% ± 0.7%, respectively. In addition, for the clinical EPID image sequences, the proposed algorithm achieves 0.32 ± 0.77 mm in the CLE and 72.1% ± 5.5% in the VOI. These results demonstrate the effectiveness of the authors' proposed method both in tumor localization and boundary tracking in EPID images. In addition, compared with two existing tracking algorithms, the proposed method achieves a higher accuracy in tumor localization.

CONCLUSIONS

In this paper, the authors presented a feasibility study of tracking tumor boundary in EPID images by using a LSM-based algorithm. Experimental results conducted on phantom and clinical EPID images demonstrated the effectiveness of the tracking algorithm for visible tumor target. Compared with previous tracking methods, the authors' algorithm has the potential to improve the tracking accuracy in radiation therapy. In addition, real-time tumor boundary information within the irradiation field will be potentially useful for further applications, such as adaptive beam delivery, dose evaluation.

Active Contour Driven by Local Gaussian Distribution Fitting and Local Signed Difference Based on Local Entropy

Intensity inhomogeneity often causes considerable difficulties in image segmentation. In order to tackle this problem, we propose a novel region-based active contour model in a variational level set formulation. We first define a data fitting energy with a local Gaussian distribution fitting (LGDF) term, which induces a local force to attract the contour and stops it at object boundaries, and a local signed difference (LSD) term based on local entropy, which possesses both local separability and global consistency. This energy is then incorporated into a level set formulation with a level set regularization term that is necessary for accurate computation in the corresponding level set method. Experimental results show that the proposed model can not only segment images with intensity inhomogeneities and weak boundaries but also be robust to the noise, initial contours.

Neuron Image Analyzer: Automated and Accurate Extraction of Neuronal Data from Low Quality Images

Image analysis software is an essential tool used in neuroscience and neural engineering to evaluate changes in neuronal structure following extracellular stimuli. Both manual and automated methods in current use are severely inadequate at detecting and quantifying changes in neuronal morphology when the images analyzed have a low signal-to-noise ratio (SNR). This inadequacy derives from the fact that these methods often include data from non-neuronal structures or artifacts by simply tracing pixels with high intensity. In this paper, we describe Neuron Image Analyzer (NIA), a novel algorithm that overcomes these inadequacies by employing Laplacian of Gaussian filter and graphical models (i.e., Hidden Markov Model, Fully Connected Chain Model) to specifically extract relational pixel information corresponding to neuronal structures (i.e., soma, neurite). As such, NIA that is based on vector representation is less likely to detect false signals (i.e., non-neuronal structures) or generate artifact signals (i.e., deformation of original structures) than current image analysis algorithms that are based on raster representation. We demonstrate that NIA enables precise quantification of neuronal processes (e.g., length and orientation of neurites) in low quality images with a significant increase in the accuracy of detecting neuronal changes post-stimulation.

CalQuo: automated, simultaneous single-cell and population-level quantification of global intracellular Ca2+ responses

Detecting intracellular calcium signaling with fluorescent calcium indicator dyes is often coupled with microscopy techniques to follow the activation state of non-excitable cells, including lymphocytes. However, the analysis of global intracellular calcium responses both at the single-cell level and in large ensembles simultaneously has yet to be automated. Here, we present a new software package, CalQuo (Calcium Quantification), which allows the automated analysis and simultaneous monitoring of global fluorescent calcium reporter-based signaling responses in up to 1000 single cells per experiment, at temporal resolutions of sub-seconds to seconds. CalQuo quantifies the number and fraction of responding cells, the temporal dependence of calcium signaling and provides global and individual calcium-reporter fluorescence intensity profiles. We demonstrate the utility of the new method by comparing the calcium-based signaling responses of genetically manipulated human lymphocytic cell lines.

Nonnegative Multiresolution Representation-Based Texture Image Classification

Effective representation of image texture is important for an image-classification task. Statistical modelling in wavelet domains has been widely used to image texture representation. However, due to the intraclass complexity and interclass diversity of textures, it is hard to use a predefined probability distribution function to fit adaptively all wavelet subband coefficients of different textures. In this article, we propose a novel modelling approach, Heterogeneous and Incrementally Generated Histogram (HIGH), to indirectly model the wavelet coefficients by use of four local features in wavelet subbands. By concatenating all the HIGHs in all wavelet subbands of a texture, we can construct a nonnegative multiresolution vector (NMV) to represent a texture image. Considering the NMV’s high dimensionality and nonnegativity, we further propose a Hessian regularized discriminative nonnegative matrix factorization to compute a low-dimensional basis of the linear subspace of NMVs. Finally, we present a texture classification approach by projecting NMVs on the low-dimensional basis. Experimental results show that our proposed texture classification method outperforms seven representative approaches.

Tracking fuzzy borders using geodesic curves with application to liver segmentation on planning CT

PURPOSE

This work aims to develop a robust and efficient method to track the fuzzy borders between liver and the abutted organs where automatic liver segmentation usually suffers, and to investigate its applications in automatic liver segmentation on noncontrast-enhanced planning computed tomography (CT) images.

METHODS

In order to track the fuzzy liver-chestwall and liver-heart borders where oversegmentation is often found, a starting point and an ending point were first identified on the coronal view images; the fuzzy border was then determined as a geodesic curve constructed by minimizing the gradient-weighted path length between these two points near the fuzzy border. The minimization of path length was numerically solved by fast-marching method. The resultant fuzzy borders were incorporated into the authors' automatic segmentation scheme, in which the liver was initially estimated by a patient-specific adaptive thresholding and then refined by a geodesic active contour model. By using planning CT images of 15 liver patients treated with stereotactic body radiation therapy, the liver contours extracted by the proposed computerized scheme were compared with those manually delineated by a radiation oncologist.

RESULTS

The proposed automatic liver segmentation method yielded an average Dice similarity coefficient of 0.930 ± 0.015, whereas it was 0.912 ± 0.020 if the fuzzy border tracking was not used. The application of fuzzy border tracking was found to significantly improve the segmentation performance. The mean liver volume obtained by the proposed method was 1727 cm(3), whereas it was 1719 cm(3) for manual-outlined volumes. The computer-generated liver volumes achieved excellent agreement with manual-outlined volumes with correlation coefficient of 0.98.

CONCLUSIONS

The proposed method was shown to provide accurate segmentation for liver in the planning CT images where contrast agent is not applied. The authors' results also clearly demonstrated that the application of tracking the fuzzy borders could significantly reduce contour leakage during active contour evolution.

Feasibility of level-set analysis of enface OCT retinal images in diabetic retinopathy

Pathology segmentation in retinal images of patients with diabetic retinopathy is important to help better understand disease processes. We propose an automated level-set method with Fourier descriptor-based shape priors. A cost function measures the difference between the current and expected output. We applied our method to enface images generated for seven retinal layers and determined correspondence of pathologies between retinal layers. We compared our method to a distance-regularized level set method and show the advantages of using well-defined shape priors. Results obtained allow us to observe pathologies across multiple layers and to obtain metrics that measure the co-localization of pathologies in different layers.

Variant of the region-scalable fitting energy for image segmentation

This paper presents a variant of the level set function based on region-scalable fitting (RSF) model for segmenting a given image into different parts. In consideration of the image local characteristics, the RSF model can efficiently and effectively segment images with intensity inhomogeneity. Instead of utilizing n level set functions to define up to 2ⁿ phases in the RSF model, our method presents a piecewise constant level set formulation for image segmentation and each phase is represented by a unique constant value. In addition, our model avoids different segmentation results caused by different initializations. The energy functional of our method is locally differentiable and convex because we do not use the nondifferentiable Heaviside and Delta functions. Comparative experiment results demonstrate that our method is much more computationally efficient. Moreover, our algorithm is robust against destructive noise.

Texture classification and retrieval using shearlets and linear regression

Statistical modeling of wavelet subbands has frequently been used for image recognition and retrieval. However, traditional wavelets are unsuitable for use with images containing distributed discontinuities, such as edges. Shearlets are a newly developed extension of wavelets that are better suited to image characterization. Here, we propose novel texture classification and retrieval methods that model adjacent shearlet subband dependences using linear regression. For texture classification, we use two energy features to represent each shearlet subband in order to overcome the limitation that subband coefficients are complex numbers. Linear regression is used to model the features of adjacent subbands; the regression residuals are then used to define the distance from a test texture to a texture class. Texture retrieval consists of two processes: the first is based on statistics in contourlet domains, while the second is performed using a pseudo-feedback mechanism based on linear regression modeling of shearlet subband dependences. Comprehensive validation experiments performed on five large texture datasets reveal that the proposed classification and retrieval methods outperform the current state-of-the-art.

Shape-based motion correction in dynamic contrast-enhanced MRI for quantitative assessment of renal function

PURPOSE

To incorporate a newly developed shape-based motion estimation scheme into magnetic resonance urography (MRU) and verify its efficacy in facilitating quantitative functional analysis.

METHODS

The authors propose a motion compensation scheme in MRU that consists of three sequential modules: MRU image acquisition, motion compensation, and quantitative functional analysis. They designed two sets of complementary experiments to evaluate the performance of the proposed method. In the first experiment, dynamic contrast enhanced (DCE) MR images were acquired from three sedated subjects, from which clinically valid estimates were derived and served as the "ground truth." Physiologically sound motion was then simulated to synthesize image sequences influenced by respiratory motion. Quantitative assessment and comparison were performed on functional estimates of Patlak number, glomerular filtration rate, and Patlak differential renal function without and with motion compensation against the ground truth. In the second experiment, the authors acquired a temporal series of noncontrast MR images under free breathing from a healthy adult subject. The performance of the proposed method on compensating real motion was evaluated by comparing the standard deviation of the obtained temporal intensity curves before and after motion compensation.

RESULTS

On DCE-MR images with simulated motion, the generated relative enhancement curves exhibited large perturbations and the Patlak numbers of the left and right kidney were significantly underestimated up to 35% and 34%, respectively, compared with the ground truth. After motion compensation, the relative enhancement curves exhibited much less perturbations and Patlak estimation errors reduced within 3% and 4% for the left and right kidneys, respectively. On clinical free-breathing MR images, the temporal intensity curves exhibited significantly reduced variations after motion compensation, with standard deviation decreased from 30.3 and 38.2 to 8.3 and 11.7 within two manually selected regions of interest, respectively.

CONCLUSIONS

The developed motion compensation method has demonstrated its ability to facilitate quantitative MRU functional analysis, with improved accuracy of pharmacokinetic modeling and quantitative parameter estimations. Future work will consider performing more intensive clinical verifications with sophisticated pharmacokinetic models and generalizing the proposed method to other quantitative DCE analysis, such as on liver or prostate function.

Feasibility of automated pancreas segmentation based on dynamic MRI

OBJECTIVE

MRI-guided radiotherapy is particularly attractive for abdominal targets with low CT contrast. To fully utilize this modality for pancreas tracking, automated segmentation tools are needed. A hybrid gradient, region growth and shape constraint (hGReS) method to segment two-dimensional (2D) upper abdominal dynamic MRI (dMRI) is developed for this purpose.

METHODS

2D coronal dynamic MR images of two healthy volunteers were acquired with a frame rate of 5 frames per second. The regions of interest (ROIs) included the liver, pancreas and stomach. The first frame was used as the source where the centres of the ROIs were manually annotated. These centre locations were propagated to the next dMRI frame. Four-neighborhood region transfer growth was performed from these initial seeds before refinement using shape constraints. RESULTS from hGReS and two other automated segmentation methods using integrated edge detection and region growth (IER) and level set, respectively, were compared with manual contours using Dice's index (DI).

RESULTS

For the first patient, the hGReS resulted in the organ segmentation accuracy as a measure by the DI (0.77) for the pancreas, superior to the level set method (0.72) and IER (0.71). The hGReS was shown to be reproducible on the second subject, achieving a DI of 0.82, 0.92 and 0.93 for the pancreas, stomach and liver, respectively. Motion trajectories derived from the hGReS were highly correlated to respiratory motion.

CONCLUSION

We have shown the feasibility of automated segmentation of the pancreas anatomy on dMRI.

ADVANCES IN KNOWLEDGE

Using the hybrid method improves segmentation robustness of low-contrast images.

Self-parameterized active contours based on regional edge structure for medical image segmentation

This work introduces a novel framework for unsupervised parameterization of region-based active contour regularization and data fidelity terms, which is applied for medical image segmentation. The work aims to relieve MDs from the laborious, time-consuming task of empirical parameterization and bolster the objectivity of the segmentation results. The proposed framework is inspired by an observed isomorphism between the eigenvalues of structure tensors and active contour parameters. Both may act as descriptors of the orientation coherence in regions containing edges. The experimental results demonstrate that the proposed framework maintains a high segmentation quality without the need of trial-and-error parameter adjustment.

A kernel-based method for markerless tumor tracking in kV fluoroscopic images

Markerless tracking of respiration-induced tumor motion in kilo-voltage (kV) fluoroscopic image sequence is still a challenging task in real time image-guided radiation therapy (IGRT). Most of existing markerless tracking methods are based on a template matching technique or its extensions that are frequently sensitive to non-rigid tumor deformation and involve expensive computation. This paper presents a kernel-based method that is capable of tracking tumor motion in kV fluoroscopic image sequence with robust performance and low computational cost. The proposed tracking system consists of the following three steps. To enhance the contrast of kV fluoroscopic image, we firstly utilize a histogram equalization to transform the intensities of original images to a wider dynamical intensity range. A tumor target in the first frame is then represented by using a histogram-based feature vector. Subsequently, the target tracking is then formulated by maximizing a Bhattacharyya coefficient that measures the similarity between the tumor target and its candidates in the subsequent frames. The numerical solution for maximizing the Bhattacharyya coefficient is performed by a mean-shift algorithm. The proposed method was evaluated by using four clinical kV fluoroscopic image sequences. For comparison, we also implement four conventional template matching-based methods and compare their performance with our proposed method in terms of the tracking accuracy and computational cost. Experimental results demonstrated that the proposed method is superior to conventional template matching-based methods.

Combining LBP difference and feature correlation for texture description

Effective characterization of texture images requires exploiting multiple visual cues from the image appearance. The local binary pattern (LBP) and its variants achieve great success in texture description. However, because the LBP(-like) feature is an index of discrete patterns rather than a numerical feature, it is difficult to combine the LBP(-like) feature with other discriminative ones by a compact descriptor. To overcome the problem derived from the nonnumerical constraint of the LBP, this paper proposes a numerical variant accordingly, named the LBP difference (LBPD). The LBPD characterizes the extent to which one LBP varies from the average local structure of an image region of interest. It is simple, rotation invariant, and computationally efficient. To achieve enhanced performance, we combine the LBPD with other discriminative cues by a covariance matrix. The proposed descriptor, termed the covariance and LBPD descriptor (COV-LBPD), is able to capture the intrinsic correlation between the LBPD and other features in a compact manner. Experimental results show that the COV-LBPD achieves promising results on publicly available data sets.

Three-dimensional prostate segmentation using level set with shape constraint based on rotational slices for 3D end-firing TRUS guided biopsy

PURPOSE

Prostate segmentation is an important step in the planning and treatment of 3D end-firing transrectal ultrasound (TRUS) guided prostate biopsy. In order to improve the accuracy and efficiency of prostate segmentation in 3D TRUS images, an improved level set method is incorporated into a rotational-slice-based 3D prostate segmentation to decrease the accumulated segmentation errors produced by the slice-by-slice segmentation method.

METHODS

A 3D image is first resliced into 2D slices in a rotational manner in both the clockwise and counterclockwise directions. All slices intersect approximately along the rotational scanning axis and have an equal angular spacing. Six to eight boundary points are selected to initialize a level set function to extract the prostate contour within the first slice. The segmented contour is then propagated to the adjacent slice and is used as the initial contour for segmentation. This process is repeated until all slices are segmented. A modified distance regularization level set method is used to segment the prostate in all resliced 2D slices. In addition, shape-constraint and local-region-based energies are imposed to discourage the evolved level set function to leak in regions with weak edges or without edges. An anchor point based energy is used to promote the level set function to pass through the initial selected boundary points. The algorithm's performance was evaluated using distance- and volume-based metrics (sensitivity (Se), Dice similarity coefficient (DSC), mean absolute surface distance (MAD), maximum absolute surface distance (MAXD), and volume difference) by comparison with expert delineations.

RESULTS

The validation results using thirty 3D patient images showed that the authors' method can obtain a DSC of 93.1% ± 1.6%, a sensitivity of 93.0% ± 2.0%, a MAD of 1.18 ± 0.36 mm, a MAXD of 3.44 ± 0.8 mm, and a volume difference of 2.6 ± 1.9 cm(3) for the entire prostate. A reproducibility experiment demonstrated that the proposed method yielded low intraobserver and interobserver variability in terms of DSC. The mean segmentation time of the authors' method for all patient 3D TRUS images was 55 ± 3.5 s, in addition to 30 ± 5 s for initialization.

CONCLUSIONS

To address the challenges involved with slice-based 3D prostate segmentation, a level set based method is proposed in this paper. This method is especially developed for a 3D end-firing TRUS guided prostate biopsy system. The extensive experimental results demonstrate that the proposed method is accurate, robust, and computationally efficient.

A shape-optimized framework for kidney segmentation in ultrasound images using NLTV denoising and DRLSE

BACKGROUND

Computer-assisted surgical navigation aims to provide surgeons with anatomical target localization and critical structure observation, where medical image processing methods such as segmentation, registration and visualization play a critical role. Percutaneous renal intervention plays an important role in several minimally-invasive surgeries of kidney, such as Percutaneous Nephrolithotomy (PCNL) and Radio-Frequency Ablation (RFA) of kidney tumors, which refers to a surgical procedure where access to a target inside the kidney by a needle puncture of the skin. Thus, kidney segmentation is a key step in developing any ultrasound-based computer-aided diagnosis systems for percutaneous renal intervention.

METHODS

In this paper, we proposed a novel framework for kidney segmentation of ultrasound (US) images combined with nonlocal total variation (NLTV) image denoising, distance regularized level set evolution (DRLSE) and shape prior. Firstly, a denoised US image was obtained by NLTV image denoising. Secondly, DRLSE was applied in the kidney segmentation to get binary image. In this case, black and white region represented the kidney and the background respectively. The last stage is that the shape prior was applied to get a shape with the smooth boundary from the kidney shape space, which was used to optimize the segmentation result of the second step. The alignment model was used occasionally to enlarge the shape space in order to increase segmentation accuracy. Experimental results on both synthetic images and US data are given to demonstrate the effectiveness and accuracy of the proposed algorithm.

RESULTS

We applied our segmentation framework on synthetic and real US images to demonstrate the better segmentation results of our method. From the qualitative results, the experiment results show that the segmentation results are much closer to the manual segmentations. The sensitivity (SN), specificity (SP) and positive predictive value (PPV) of our segmentation result can reach 95%, 96% and 91% respectively; As well as we compared our results with the edge-based level set and level set with shape prior method by means of the same quantitative index, such as SN, SP, PPV, which have corresponding values of 97%, 88%, 78% and 81%, 91%, 80% respectively.

CONCLUSIONS

We have found NLTV denosing method is a good initial process for the ultrasound segmentation. This initial process can make us use simple segmentation method to get satisfied results. Furthermore, we can get the final segmentation results with smooth boundary by using the shape prior after the segmentation process. Every step enjoy simple energy model and every step in this framework is needed to keep a good robust and convergence property.

A quantitative study of motion estimation methods on 4D cardiac gated SPECT reconstruction

PURPOSE

Motion-compensated temporal processing can have a major impact on improving the image quality in gated cardiac single photon emission computed tomography (SPECT). In this work, we investigate the effect of different optical flow estimation methods for motion-compensated temporal processing in gated SPECT. In particular, we explore whether better motion estimation can substantially improve reconstructed image quality, and how the estimated motion would compare to the ideal case of known motion in terms of reconstruction.

METHODS

We consider the following three methods for obtaining the image motion in 4D reconstruction: (1) the Horn-Schunck optical flow equation (OFE) method, (2) a recently developed periodic OFE method, and (3) known cardiac motion derived from the NURBS-based cardiac-torso (NCAT) phantom. The periodic OFE method is used to exploit the inherent periodic nature in cardiac gated images. In this method, the optical flow in a sequence is modeled by a Fourier harmonic representation, which is then estimated from the image data. We study the impact of temporal processing on 4D reconstructions when the image motion is obtained with the different methods above. For quantitative evaluation, we use simulated imaging with multiple noise realizations from the NCAT phantom, where different patient geometry and lesion sizes are also considered. To quantify the reconstruction results, we use the following measures of reconstruction accuracy and defect detection in the myocardium: (1) overall error level in the myocardium, (2) regional accuracy of the left ventricle (LV) wall, (3) accuracy of regional time activity curves of the LV, and (4) perfusion defect detectability with a channelized Hotelling observer (CHO). In addition, we also examine the effect of noise on the distortion in the reconstructed LV wall shape by detecting its contours. As a preliminary demonstration, these methods are also tested on two sets of clinical acquisitions.

RESULTS

For the different quantitative measures considered, the periodic OFE further improved the reconstruction accuracy of the myocardium compared to OFE in 4D reconstruction; its improvement in reconstruction almost matched that of the known motion. Specifically, the overall mean-squared error in the myocardium was reduced by over 20% with periodic OFE; with noise level fixed at 10%, the regional bias on the LV was reduced from 20% (OFE) to 14% (periodic OFE), compared to 11% by the known motion. In addition, the CHO results show that there was also improvement in lesion detectability with the periodic OFE. The regional time activity curves obtained with the periodic OFE were also observed to be more consistent with the reference; in addition, the contours of the reconstructed LV wall with the periodic OFE were demonstrated to show less degree of variations among different noise realizations. Such improvements were also consistent with the results obtained from the clinical acquisitions.

CONCLUSIONS

Use of improved optical flow estimation can further improve the accuracy of reconstructed images in 4D. The periodic OFE method not only can achieve improvements over the traditional OFE, but also can almost match that of the known motion in terms of the several quality measures considered.