Random walks for image segmentation

A Continuous Random Walk Model With Explicit Coherence Regularization for Image Segmentation

Random walk is a popular and efficient algorithm for image segmentation, especially for extracting regions of interest (ROIs). One difficulty with the random walk algorithm is the requirement for solving a huge sparse linear system when applied to large images. Another limitation is its sensitivity to seeds distribution, i.e., the segmentation result depends on the number of seeds as well as their placement, which puts a burden on users. In this paper, we first propose a continuous random walk model with explicit coherence regularization (CRWCR) for the extracted ROI, which helps to reduce the seeds sensitivity, so as to reduce the user interactions. Then, a very efficient algorithm to solve the CRWCR model will be developed, which helps to remove the difficulty of solving huge linear systems. Our algorithm consists of two stages: initialization by performing one-dimensional random walk sweeping based on user-provided seeds, followed by the alternating direction scheme, i.e., Peaceman-Rachford scheme for further correction. The first stage aims to provide a good initial guess for the ROI, and it is very fast since we just solve a limited number of one-dimensional random walk problems. Then, this initial guess is evolved to the ideal solution by applying the second stage, which should also be very efficient since it fits well for GPU computing, and 10 iterations are usually sufficient for convergence. Numerical experiments are provided to validate the proposed model as well as the efficiency of the two-stage algorithm.

Automatic Detection of B-Lines in In Vivo Lung Ultrasound

This paper proposes an automatic method for accurate detection and visualization of B-lines in ultrasound lung scans, which provides a quantitative measure for the number of B-lines present. All the scans used in this study were acquired using a BK3000 ultrasound scanner (BK Ultrasound, Herlev, Denmark) driving a 5.5-MHz linear transducer (BK Ultrasound). Four healthy subjects and four patients, after major surgery with pulmonary edema, were scanned at four locations on each lung for B-line examination. Eight sequences of 50 frames were acquired for each subject yielding 64 sequences in total. The proposed algorithm was applied to all 3200 in-vivo lung ultrasound images. The results showed that the average number of B-lines was 0.28±0.06 (Mean±Std) in scans belonging to the patients compared to 0.03 ± 0.06 (Mean ± Std) in the healthy subjects. Also, the Mann-Whitney test showed a significant difference between the two groups with the p -value of 0.015, and indicating that the proposed algorithm was able to differentiate between the healthy volunteers and the patients. In conclusion, the method can be used to automatically and to quantitatively characterize the distribution of B-lines for diagnosing pulmonary edema.

Robust Pulmonary Nodule Segmentation in CT Image for Juxta-pleural and Juxta-vascular Case

Background:

Lung cancer is a greatest threat to people's health and life. CT image leads to unclear boundary segmentation. Segmentation of irregular nodules and complex structure, boundary information is not well considered and lung nodules have always been a hot topic.

Objective:

In this study, the pulmonary nodule segmentation is accomplished with the new graph cut algorithm. The problem of segmenting the juxta-pleural and juxta-vascular nodules was investigated which is based on graph cut algorithm.

Methods:

Firstly, the inflection points by the curvature was decided. Secondly, we used kernel graph cut to segment the nodules for the initial edge. Thirdly, the seeds points based on cast raying method is performed; lastly, a novel geodesic distance function is proposed to improve the graph cut algorithm and applied in lung nodules segmentation.

Results:

The new algorithm has been tested on total 258 nodules. Table 1 summarizes the morphologic features of all the nodules and given the results between the successful segmentation group and the poor/failed segmentation group. Figure 1 to Fig. (12) shows segmentation effect of Juxta-vascular nodules, Juxta-pleural nodules, and comparted with the other interactive segmentation methods.

Conclusion:

The experimental verification shows better results with our algorithm, the results will measure the volume numerical approach to nodule volume. The results of lung nodules segmentation in this study are as good as the results obtained by the other methods.

Segmentation of lung parenchyma in CT images using CNN trained with the clustering algorithm generated dataset

Background

Lung segmentation constitutes a critical procedure for any clinical-decision supporting system aimed to improve the early diagnosis and treatment of lung diseases. Abnormal lungs mainly include lung parenchyma with commonalities on CT images across subjects, diseases and CT scanners, and lung lesions presenting various appearances. Segmentation of lung parenchyma can help locate and analyze the neighboring lesions, but is not well studied in the framework of machine learning.

Methods

We proposed to segment lung parenchyma using a convolutional neural network (CNN) model. To reduce the workload of manually preparing the dataset for training the CNN, one clustering algorithm based method is proposed firstly. Specifically, after splitting CT slices into image patches, the k-means clustering algorithm with two categories is performed twice using the mean and minimum intensity of image patch, respectively. A cross-shaped verification, a volume intersection, a connected component analysis and a patch expansion are followed to generate final dataset. Secondly, we design a CNN architecture consisting of only one convolutional layer with six kernels, followed by one maximum pooling layer and two fully connected layers. Using the generated dataset, a variety of CNN models are trained and optimized, and their performances are evaluated by eightfold cross-validation. A separate validation experiment is further conducted using a dataset of 201 subjects (4.62 billion patches) with lung cancer or chronic obstructive pulmonary disease, scanned by CT or PET/CT. The segmentation results by our method are compared with those yielded by manual segmentation and some available methods.

Results

A total of 121,728 patches are generated to train and validate the CNN models. After the parameter optimization, our CNN model achieves an average F-score of 0.9917 and an area of curve up to 0.9991 for classification of lung parenchyma and non-lung-parenchyma. The obtain model can segment the lung parenchyma accurately for 201 subjects with heterogeneous lung diseases and CT scanners. The overlap ratio between the manual segmentation and the one by our method reaches 0.96.

Conclusions

The results demonstrated that the proposed clustering algorithm based method can generate the training dataset for CNN models. The obtained CNN model can segment lung parenchyma with very satisfactory performance and have the potential to locate and analyze lung lesions.

Macroscopic defects upon decomposition of CO2 clathrate hydrate crystals

Cracks and decomposition barriers observed in time-lapse micro-computed tomography measurements challenge existing models of gas hydrate decomposition.

Probabilistic Diffusion for Interactive Image Segmentation

This paper presents an interactive image segmentation approach in which we formulate segmentation as a probabilistic estimation problem based on the prior user intention. Instead of directly measuring the relationship between pixels and labels, we first estimate the distances between pixel pairs and label pairs using a probabilistic framework. Then, binary probabilities with label pairs are naturally converted to unary probabilities with labels. The higher order relationship helps improve the robustness to user inputs. To improve segmentation accuracy, a likelihood learning framework is proposed to fuse the region and the boundary information of the image by imposing a smoothing constraint on the unary potentials. Furthermore, we establish an equivalence relationship between likelihood learning and likelihood diffusion and propose an iterative diffusion-based optimization strategy to maintain computational efficiency. Experiments on the Berkeley segmentation data set and Microsoft GrabCut database demonstrate that the proposed method can obtain better performance than the state-of-the-art methods.

Tumor co-segmentation in PET/CT using multi-modality fully convolutional neural network

Automatic tumor segmentation from medical images is an important step for computer-aided cancer diagnosis and treatment. Recently, deep learning has been successfully applied to this task, leading to state-of-the-art performance. However, most of existing deep learning segmentation methods only work for a single imaging modality. PET/CT scanner is nowadays widely used in the clinic, and is able to provide both metabolic information and anatomical information through integrating PET and CT into the same utility. In this study, we proposed a novel multi-modality segmentation method based on a 3D fully convolutional neural network (FCN), which is capable of taking account of both PET and CT information simultaneously for tumor segmentation. The network started with a multi-task training module, in which two parallel sub-segmentation architectures constructed using deep convolutional neural networks (CNNs) were designed to automatically extract feature maps from PET and CT respectively. A feature fusion module was subsequently designed based on cascaded convolutional blocks, which re-extracted features from PET/CT feature maps using a weighted cross entropy minimization strategy. The tumor mask was obtained as the output at the end of the network using a softmax function. The effectiveness of the proposed method was validated on a clinic PET/CT dataset of 84 patients with lung cancer. The results demonstrated that the proposed network was effective, fast and robust and achieved significantly performance gain over CNN-based methods and traditional methods using PET or CT only, two V-net based co-segmentation methods, two variational co-segmentation methods based on fuzzy set theory and a deep learning co-segmentation method using W-net.

Infrared Small Target Detection via Modified Random Walks

Infrared small target detection under intricate background and heavy noise is one of the crucial tasks in the field of remote sensing. Conventional algorithms can fail in detecting small targets due to the low signal-to-noise ratios of the images. To solve this problem, an effective infrared small target detection algorithm inspired by random walks is presented in this paper. The novelty of our contribution involves the combination of the local contrast feature and the global uniqueness of the small targets. Firstly, the original pixel-wise image is transformed into an multi-dimensional image with respect to the local contrast measure. Secondly, a reconstructed seeds selection map (SSM) is generated based on the multi-dimensional image. Then, an adaptive seeds selection method is proposed to automatically select the foreground seeds potentially placed in the areas of the small targets in the SSM. After that, a confidence map is constructed using a modified random walks (MRW) algorithm to represent the global uniqueness of the small targets. Finally, we segment the targets from the confidence map by utilizing an adaptive threshold. Extensive experimental evaluation results on a real test dataset demonstrate that our algorithm is superior to the state-of-the-art algorithms in both target enhancement and detection performance.

Molecular PET imaging in adaptive radiotherapy: brain

INTRODUCTION

Owing to their heterogeneity and radioresistance, the prognosis of primitive brain tumors, which are mainly glial tumors, remains poor. Dose escalation in radioresistant areas is a potential issue for improving local control and overall survival. This review focuses on advances in biological and metabolic imaging of brain tumors that are proving to be essential for defining tumor target volumes in radiation therapy (RT) and for increasing the use of DPRT (dose painting RT) and ART (adaptative RT), to optimize dose in radio-resistant areas.

EVIDENCE ACQUISITION

Various biological imaging modalities such as PET (hypoxia, glucidic metabolism, protidic metabolism, cellular proliferation, inflammation, cellular membrane synthesis) and MRI (spectroscopy) may be used to identify these areas of radioresistance. The integration of these biological imaging modalities improves the diagnosis, prognosis and treatment of brain tumors.

EVIDENCE SYNTHESIS

Technological improvements (PET and MRI), the development of research, and intensive cooperation between different departments are necessary before using daily metabolic imaging (PET and MRI) to treat patients with brain tumors.

CONCLUSIONS

The adaptation of treatment volumes during RT (ART) seems promising, but its development requires improvements in several areas and an interdisciplinary approach involving radiology, nuclear medicine and radiotherapy. We review the literature on biological imaging to outline the perspectives for using DPRT and ART in brain tumors.

Modified Superpixel Segmentation for Digital Surface Model Refinement and Building Extraction from Satellite Stereo Imagery

Superpixels, as a state-of-the-art segmentation paradigm, have recently been widely used in computer vision and pattern recognition. Despite the effectiveness of these algorithms, there are still many limitations and challenges dealing with Very High-Resolution (VHR) satellite images especially in complex urban scenes. In this paper, we develop a superpixel algorithm as a modified edge-based version of Simple Linear Iterative Clustering (SLIC), which is here called ESLIC, compatible with VHR satellite images. Then, based on the modified properties of generated superpixels, a heuristic multi-scale approach for building extraction is proposed, based on the stereo satellite imagery along with the corresponding Digital Surface Model (DSM). First, to generate the modified superpixels, an edge-preserving term is applied to retain the main building boundaries and edges. The resulting superpixels are then used to initially refine the stereo-extracted DSM. After shadow and vegetation removal, a rough building mask is obtained from the normalized DSM, which highlights the appropriate regions in the image, to be used as the input of a multi-scale superpixel segmentation of the proper areas to determine the superpixels inside the building. Finally, these building superpixels with different scales are integrated and the output is a unified building mask. We have tested our methods on building samples from a WorldView-2 dataset. The results are promising, and the experiments show that superpixels generated with the proposed ESLIC algorithm are more adherent to the building boundaries, and the resulting building mask retains urban object shape better than those generated with the original SLIC algorithm.

Nanomagnetic properties of the meteorite cloudy zone

Meteorites contain a record of their thermal and magnetic history, written in the intergrowths of iron-rich and nickel-rich phases that formed during slow cooling. Of intense interest from a magnetic perspective is the "cloudy zone," a nanoscale intergrowth containing tetrataenite-a naturally occurring hard ferromagnetic mineral that has potential applications as a sustainable alternative to rare-earth permanent magnets. Here we use a combination of high-resolution electron diffraction, electron tomography, atom probe tomography (APT), and micromagnetic simulations to reveal the 3D architecture of the cloudy zone with subnanometer spatial resolution and model the mechanism of remanence acquisition during slow cooling on the meteorite parent body. Isolated islands of tetrataenite are embedded in a matrix of an ordered superstructure. The islands are arranged in clusters of three crystallographic variants, which control how magnetic information is encoded into the nanostructure. The cloudy zone acquires paleomagnetic remanence via a sequence of magnetic domain state transformations (vortex to two domain to single domain), driven by Fe-Ni ordering at 320 °C. Rather than remanence being recorded at different times at different positions throughout the cloudy zone, each subregion of the cloudy zone records a coherent snapshot of the magnetic field that was present at 320 °C. Only the coarse and intermediate regions of the cloudy zone are found to be suitable for paleomagnetic applications. The fine regions, on the other hand, have properties similar to those of rare-earth permanent magnets, providing potential routes to synthetic tetrataenite-based magnetic materials.

Differentiation of triple-negative breast cancer from other subtypes through whole-tumor histogram analysis on multiparametric MR imaging

PURPOSE

To identify triple-negative (TN) breast cancer imaging biomarkers in comparison to other molecular subtypes using multiparametric MR imaging maps and whole-tumor histogram analysis.

MATERIALS AND METHODS

This retrospective study included 134 patients with invasive ductal carcinoma. Whole-tumor histogram-based texture features were extracted from a quantitative ADC map and DCE semi-quantitative maps (washin and washout). Univariate analysis using the Student's t test or Mann-Whitney U test was performed to identify significant variables for differentiating TN cancer from other subtypes. The ROC curves were generated based on the significant variables identified from the univariate analysis. The AUC, sensitivity, and specificity for subtype differentiation were reported.

RESULTS

The significant parameters on the univariate analysis achieved an AUC of 0.710 (95% confidence interval [CI] 0.562, 0.858) with a sensitivity of 63.6% and a specificity of 73.1% at the best cutoff point for differentiating TN cancers from Luminal A cancers. An AUC of 0.763 (95% CI 0.608, 0.917) with a sensitivity of 86.4% and a specificity of 72.2% was achieved for differentiating TN cancers from human epidermal growth factor receptor 2 (HER2) positive cancers. Also, an AUC of 0.683 (95% CI 0.556, 0.809) with a sensitivity of 54.5% and a specificity of 83.9% was achieved for differentiating TN cancers from non-TN cancers. There was no significant feature on the univariate analysis for TN cancers versus Luminal B cancers.

CONCLUSIONS

Whole-tumor histogram-based imaging features derived from ADC, along with washin and washout maps, provide a non-invasive analytical approach for discriminating TN cancers from other subtypes.

KEY POINTS

• Whole-tumor histogram-based features on MR multiparametric maps can help to assess biological characterization of breast cancer. • Histogram-based texture analysis may predict the molecular subtypes of breast cancer. • Combined DWI and DCE evaluation helps to identify triple-negative breast cancer.

Volumetric assessment of spontaneous mechanical activities by simultaneous multi-slice MRI techniques with correlation to muscle fiber orientation

The purpose of this work was assessment of volumetric characteristics of spontaneous mechanical activities in musculature (SMAMs) by diffusion-weighted simultaneous multi-slice (DW-SMS) imaging and spatial correlation to anatomical structure, as revealed by fusion to fiber tractographic information derived from diffusion-tensor imaging (DTI). The feasibility of using DW-SMS to image spontaneous events in human musculature was assessed by phantom measurements. Series of DW-SMS images and DTI datasets were recorded from the resting calf of three human subjects. Simultaneously recorded SMAMs in multiple slices were analyzed regarding spatial extension by the Kolmogorov-Smirnov test. Direct correlation of spatial distribution of SMAMs and fiber orientation was investigated by mapping of muscle fibers to multi-slice SMAM datasets. The DW-SMS strategy allows simultaneous assessment of SMAMs in several slices of resting skeletal musculature, since 73.9% of SMAM-affected volumes have shown SMAMs in multiple DW-SMS slices. Spatial extension of SMAMs was highly correlated over different simultaneously recorded DW-SMS slices, and affected areas followed the orientation of muscle fibers with a connectivity ratio up to 57.18 ± 14.80% based on event count and connectivity count maps. In 89.2% of all SMAM-affected datasets muscle fiber connectivity was shown in at least two adjacent slices. Direct correlation between SMAMs in human lower leg musculature and underlying anatomical structure was revealed by high muscle fiber connectivity (89.2%). SMAMs have shown a wide distribution along the longitudinal muscle direction (73.9% in multiple DW-SMS slices) with direct involvement of muscle fibers. Correlation between SMAMs in multiple DW-SMS slices and crossing muscular fiber tracts provides evidence that SMAMs result from physiological processes in musculature. Fusion of DW-SMS with DTI facilitates non-invasive studies of muscle fiber involvement in SMAMs in resting muscle.

Unsupervised machine learning of radiomic features for predicting treatment response and overall survival of early stage non-small cell lung cancer patients treated with stereotactic body radiation therapy

BACKGROUND AND PURPOSE

To predict treatment response and survival of NSCLC patients receiving stereotactic body radiation therapy (SBRT), we develop an unsupervised machine learning method for stratifying patients and extracting meta-features simultaneously based on imaging data.

MATERIAL AND METHODS

This study was performed based on an 18F-FDG-PET dataset of 100 consecutive patients who were treated with SBRT for early stage NSCLC. Each patient's tumor was characterized by 722 radiomic features. An unsupervised two-way clustering method was used to identify groups of patients and radiomic features simultaneously. The groups of patients were compared in terms of survival and freedom from nodal failure. Meta-features were computed for building survival models to predict survival and free of nodal failure.

RESULTS

Differences were found between 2 groups of patients when the patients were clustered into 3 groups in terms of both survival (p = 0.003) and freedom from nodal failure (p = 0.038). Average concordance measures for predicting survival and nodal failure were 0.640±0.029 and 0.664±0.063 respectively, better than those obtained by prediction models built upon clinical variables (p < 0.04).

CONCLUSIONS

The evaluation results demonstrate that our method allows us to stratify patients and predict survival and freedom from nodal failure with better performance than current alternative methods.

The trapping problem of the weighted scale-free treelike networks for two kinds of biased walks

It has been recently reported that trapping problem can characterize various dynamical processes taking place on complex networks. However, most works focused on the case of binary networks, and dynamical processes on weighted networks are poorly understood. In this paper, we study two kinds of biased walks including standard weight-dependent walk and mixed weight-dependent walk on the weighted scale-free treelike networks with a trap at the central node. Mixed weight-dependent walk including non-nearest neighbor jump appears in many real situations, but related studies are much less. By the construction of studied networks in this paper, we determine all the eigenvalues of the fundamental matrix for two kinds of biased walks and show that the largest eigenvalue has an identical dominant scaling as that of the average trapping time (ATT). Thus, we can obtain the leading scaling of ATT by a more convenient method and avoid the tedious calculation. The obtained results show that the weight factor has a significant effect on the ATT, and the smaller the value of the weight factor, the more efficient the trapping process is. Comparing the standard weight-dependent walk with mixed weight-dependent walk, although next-nearest-neighbor jumps have no main effect on the trapping process, they can modify the coefficient of the dominant term for the ATT.

A collaborative computer aided diagnosis (C-CAD) system with eye-tracking, sparse attentional model, and deep learning

Computer aided diagnosis (CAD) tools help radiologists to reduce diagnostic errors such as missing tumors and misdiagnosis. Vision researchers have been analyzing behaviors of radiologists during screening to understand how and why they miss tumors or misdiagnose. In this regard, eye-trackers have been instrumental in understanding visual search processes of radiologists. However, most relevant studies in this aspect are not compatible with realistic radiology reading rooms. In this study, we aim to develop a paradigm shifting CAD system, called collaborative CAD (C-CAD), that unifies CAD and eye-tracking systems in realistic radiology room settings. We first developed an eye-tracking interface providing radiologists with a real radiology reading room experience. Second, we propose a novel algorithm that unifies eye-tracking data and a CAD system. Specifically, we present a new graph based clustering and sparsification algorithm to transform eye-tracking data (gaze) into a graph model to interpret gaze patterns quantitatively and qualitatively. The proposed C-CAD collaborates with radiologists via eye-tracking technology and helps them to improve their diagnostic decisions. The C-CAD uses radiologists' search efficiency by processing their gaze patterns. Furthermore, the C-CAD incorporates a deep learning algorithm in a newly designed multi-task learning platform to segment and diagnose suspicious areas simultaneously. The proposed C-CAD system has been tested in a lung cancer screening experiment with multiple radiologists, reading low dose chest CTs. Promising results support the efficiency, accuracy and applicability of the proposed C-CAD system in a real radiology room setting. We have also shown that our framework is generalizable to more complex applications such as prostate cancer screening with multi-parametric magnetic resonance imaging (mp-MRI).

Hybrid method combining superpixel, random walk and active contour model for fast and accurate liver segmentation

Organ segmentation is an important pre-processing step in surgery planning and computer-aided diagnosis. In this paper, we propose a fast and accurate liver segmentation framework. Our proposed method combines a knowledge-based slice-by-slice Random Walk (RW) segmentation algorithm (proposed in our previous work) with a superpixel algorithm called the Contrast-enhanced Compact Watershed (CCWS) method to reduce computing time and memory costs. Compared to the commonly used Simple Linear Iterative Clustering (SLIC), we demonstrate that our CCWS is more appropriate for liver segmentation. To improve the methods accuracy, we use a modified narrow band active contour model as a refinement after the initial segmentation. The experiments showed that the superpixel-based slice-by-slice RW could segment the entire liver with improved speed, and the modified active contour model is more precise than the original Chan-Vese Model. As a result, the proposed framework is able to quickly and accurately segment the entire liver.

Maximal entropy random walk on heterogenous network for MIRNA-disease Association prediction

The last few decades have verified the vital roles of microRNAs in the development of human diseases and witnessed the increasing interest in the prediction of potential disease-miRNA associations. Owning to the open access of many miRNA-related databases, up until recently, kinds of feasible in silico models have been proposed. In this work, we developed a computational model of Maximal Entropy Random Walk on heterogenous network for MiRNA-disease Association prediction (MERWMDA). MERWMDA integrated known disease-miRNA association, pair-wise functional relation of miRNAs and pair-wise semantic relation of diseases into a heterogenous network comprised of disease and miRNA nodes full of information. As a kind of widely-applied biased walk process with more randomness, MERW was then implemented on the heterogenous network to reveal potential disease-miRNA associations. Cross validation was further performed to evaluate the performance of MERWMDA. As a result, MERWMDA obtained AUCs of 0.8966 and 0.8491 respectively in the aspect of global and local leave-one-out cross validation. What' more, three different case study strategies on four human complex diseases were conducted to comprehensively assess the quality of the model. Specifically, one kind of case study on Esophageal cancer and Prostate cancer were conducted based on HMDD v2.0 database. 94% and 88% out of the top 50 ranked miRNAs were confirmed by recent literature, respectively. To simulate new disease without known related miRNAs, Lung cancer (confirmed ratio 94%) associated miRNAs were removed for case study. Lymphoma (verified ratio 88%) was adopted to assess the prediction robustness of MERWMDA based on HMDD v1.0 database. We anticipated that MERWMDA could offer valuable candidates for in vitro biomedical experiments in future.

Exploring Weakly Labeled Images for Video Object Segmentation With Submodular Proposal Selection

Video object segmentation (VOS) is important for various computer vision problems, and handling it with minimal human supervision is highly desired for the large-scale applications. To bring down the supervision, existing approaches largely follow a data mining perspective by assuming the availability of multiple videos sharing the same object categories. It, however, would be problematic for the tasks that consume a single video. To address this problem, this paper proposes a novel approach that explores weakly labeled images to solve video object segmentation. Given a video labeled with a target category, images labeled with the same category are collected, from which noisy object exemplars are automatically discovered. After that the proposed approach extracts a set of region proposals on various frames and efficiently matches them with massive noisy exemplars in terms of appearance and spatial context. We then jointly select the best proposals across the video by solving a novel submodular problem that combines region voting and global region matching. Finally, the localization results are leveraged as strong supervision to guide pixel-level segmentation. Extensive experiments are conducted on two challenging public databases: Youtube-Objects and DAVIS. The results suggest that the proposed approach improves over previous weakly supervised/unsupervised approaches significantly, showing a performance even comparable with the several approaches supervised by the costly manual segmentations.

Unsupervised Change Detection Using Fast Fuzzy Clustering for Landslide Mapping from Very High-Resolution Images

Change detection approaches based on image segmentation are often used for landslide mapping (LM) from very high-resolution (VHR) remote sensing images. However, these approaches usually have two limitations. One is that they are sensitive to thresholds used for image segmentation and require too many parameters. The other one is that the computational complexity of these approaches depends on the image size, and thus they require a long execution time for very high-resolution (VHR) remote sensing images. In this paper, an unsupervised change detection using fast fuzzy c-means clustering (CDFFCM) for LM is proposed. The proposed CDFFCM has two contributions. The first is that we employ a Gaussian pyramid-based fast fuzzy c-means (FCM) clustering algorithm to obtain candidate landslide regions that have a better visual effect due to the utilization of image spatial information. The second is that we use the difference of image structure information instead of grayscale difference to obtain more accurate landslide regions. Three comparative approaches, edge-based level-set (ELSE), region-based level-set (RLSE), and change detection-based Markov random field (CDMRF), and the proposed CDFFCM are evaluated in three true landslide cases in the Lantau area of Hong Kong. The experiments show that the proposed CDFFCM is superior to three comparative approaches in terms of higher accuracy, fewer parameters, and shorter execution time.

Anatomical Structure Segmentation in Ultrasound Volumes Using Cross Frame Belief Propagating Iterative Random Walks

Ultrasound (US) is widely used as a low-cost alternative to computed tomography or magnetic resonance and primarily for preliminary imaging. Since speckle intensity in US images is inherently stochastic, readers are often challenged in their ability to identify the pathological regions in a volume of a large number of images. This paper introduces a generalized approach for volumetric segmentation of structures in US images and volumes. We employ an iterative random walks (IRW) solver, a random forest learning model, and a gradient vector flow (GVF) based interframe belief propagation technique for achieving cross-frame volumetric segmentation. At the start, a weak estimate of the tissue structure is obtained using estimates of parameters of a statistical mechanics model of US tissue interaction. Ensemble learning of these parameters further using a random forest is used to initialize the segmentation pipeline. IRW is used for correcting the contour in various steps of the algorithm. Subsequently, a GVF-based interframe belief propagation is applied to adjacent frames based on the initialization of contour using information in the current frame to segment the complete volume by frame-wise processing. We have experimentally evaluated our approach using two different datasets. Intravascular ultrasound (IVUS) segmentation was evaluated using 10 pullbacks acquired at 20 MHz and thyroid US segmentation is evaluated on 16 volumes acquired at [Formula: see text] MHz. Our approach obtains a Jaccard score of [Formula: see text] for IVUS segmentation and [Formula: see text] for thyroid segmentation while processing each frame in [Formula: see text] for the IVUS and in [Formula: see text] for thyroid segmentation without the need of any computing accelerators such as GPUs.

Locally advanced gastric cancer: total iodine uptake to predict the response of primary lesion to neoadjuvant chemotherapy

PURPOSE

Pathologic response to neoadjuvant chemotherapy is a prognostic factor in many cancer types. However, the existing evaluative criteria are deficient. We sought to prospectively evaluate the total iodine uptake derived from dual-energy computed tomography (DECT) in predicting treatment efficacy and progression-free survival (PFS) time in gastric cancer after neoadjuvant chemotherapy.

METHODS

From October 2012 to December 2015, 44 patients with locally advanced gastric cancer were examined with DECT 1 week before and three cycles after neoadjuvant chemotherapy. The percentage changes in tumor area (%ΔS), diameter (%ΔD), and density (%ΔHU) were calculated to evaluate the WHO, RESCIST, and Choi criteria. The percentage changes in tumor volume (%ΔV) and total iodine uptake of portal phase (%ΔTIU-p) were also calculated to determine cut-off values by ROC curves. The correlation between the different criteria and histopathologic tumor regression grade (Becker score) or PFS were statistically analyzed.

RESULTS

Forty-four patients were divided into responders and non-responders according to 43.34% volume reduction (P = 0.002) and 63.87% (P = 0.002) TIU-p reduction, respectively. The %ΔTIU-p showed strong (r = 0.602, P = 0.000) and %ΔV showed moderate (r = 0.416, P = 0.005), while the WHO (r = 0.075, P = 0.627), RECIST (r = 0.270, P = 0.077) and Choi criteria (r = 0.238, P = 0.120) showed no correlation with the Becker score. The differences in PFS time between the responder and non-responder groups were significant according to %ΔTIU-p and Choi criteria (P = 0.001 and P = 0.013, respectively).

CONCLUSIONS

The TIU-p can help predict pathological regression in advanced gastric cancer patients after neoadjuvant chemotherapy. In addition, the %ΔTIU-p could be one of the potentially valuable predictive parameters of the PFS time.

The dynamics of knowledge acquisition via self-learning in complex networks

Studies regarding knowledge organization and acquisition are of great importance to understand areas related to science and technology. A common way to model the relationship between different concepts is through complex networks. In such representations, networks' nodes store knowledge and edges represent their relationships. Several studies that considered this type of structure and knowledge acquisition dynamics employed one or more agents to discover node concepts by walking on the network. In this study, we investigate a different type of dynamics adopting a single node as the "network brain." Such a brain represents a range of real systems such as the information about the environment that is acquired by a person and is stored in the brain. To store the discovered information in a specific node, the agents walk on the network and return to the brain. We propose three different dynamics and test them on several network models and on a real system, which is formed by journal articles and their respective citations. The results revealed that, according to the adopted walking models, the efficiency of self-knowledge acquisition has only a weak dependency on topology and search strategy.

Image Segmentation for Intensity Inhomogeneity in Presence of High Noise

Automated segmentation of fine objects details in a given image is becoming of crucial interest in different imaging fields. In this paper, we propose a new variational level-set model for both global and interactive\selective segmentation tasks, which can deal with intensity inhomogeneity and the presence of noise. The proposed method maintains the same performance on clean and noisy vector-valued images. The model utilizes a combination of locally computed denoising constrained surface and a denoising fidelity term to ensure a fine segmentation of local and global features of a given image. A two-phase level-set formulation has been extended to a multi-phase formulation to successfully segment medical images of the human brain. Comparative experiments with state-of-the-art models show the advantages of the proposed method.

3D Randomized Connection Network with Graph-based Label Inference

In this paper, a novel 3D deep learning network is proposed for brain MR image segmentation with randomized connection, which can decrease the dependency between layers and increase the network capacity. The convolutional LSTM and 3D convolution are employed as network units to capture the long-term and short-term 3D properties respectively. To assemble these two kinds of spatial-temporal information and refine the deep learning outcomes, we further introduce an efficient graph-based node selection and label inference method. Experiments have been carried out on two publicly available databases and results demonstrate that the proposed method can obtain competitive performances as compared with other state-of-the-art methods.

Automatic aortic valve landmark localization in coronary CT angiography using colonial walk

The minimally invasive transcatheter aortic valve implantation (TAVI) is the most prevalent method to treat aortic valve stenosis. For pre-operative surgical planning, contrast-enhanced coronary CT angiography (CCTA) is used as the imaging technique to acquire 3-D measurements of the valve. Accurate localization of the eight aortic valve landmarks in CT images plays a vital role in the TAVI workflow because a small error risks blocking the coronary circulation. In order to examine the valve and mark the landmarks, physicians prefer a view parallel to the hinge plane, instead of using the conventional axial, coronal or sagittal view. However, customizing the view is a difficult and time-consuming task because of unclear aorta pose and different artifacts of CCTA. Therefore, automatic localization of landmarks can serve as a useful guide to the physicians customizing the viewpoint. In this paper, we present an automatic method to localize the aortic valve landmarks using colonial walk, a regression tree-based machine-learning algorithm. For efficient learning from the training set, we propose a two-phase optimized search space learning model in which a representative point inside the valvular area is first learned from the whole CT volume. All eight landmarks are then learned from a smaller area around that point. Experiment with preprocedural CCTA images of TAVI undergoing patients showed that our method is robust under high stenotic variation and notably efficient, as it requires only 12 milliseconds to localize all eight landmarks, as tested on a 3.60 GHz single-core CPU.

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).

Myocardium segmentation from DE MRI with guided random walks and sparse shape representation

PURPOSE

For patients with myocardial infarction (MI), delayed enhancement (DE) cardiovascular magnetic resonance imaging (MRI) is a sensitive and well-validated technique for the detection and visualization of MI. The myocardium viability assessment with DE MRI is important in diagnosis and treatment management, where myocardium segmentation is a prerequisite. However, few academic works have focused on automated myocardium segmentation from DE images. In this study, we aim to develop an automatic myocardium segmentation algorithm that targets DE images.

METHODS

We propose a segmentation framework based on both prior shape knowledge and image intensity. Instead of the strong request of the pre-segmentation of cine MRI in the same session, we use the sparse representation method to model the myocardium shape. Data from the Cardiac MR Left Ventricle Segmentation Challenge (2009) are used to build the shape template repository. The method of guided random walks is used to integrate the shape model and intensity information. An iterative approach is used to gradually improve the results.

RESULTS

The proposed method was tested on the DE MRI data from 30 MI patients. The proposed method achieved Dice similarity coefficients (DSC) of 74.60 ± 7.79% with 201 shape templates and 73.56 ± 6.32% with 56 shape templates, which were close to the inter-observer difference (73.94 ± 5.12%). To test the generalization of the proposed method to routine clinical images, the DE images of 10 successive new patients were collected, which were unseen during the method development and parameter tuning, and a DSC of 76.02 ± 7.43% was achieved.

CONCLUSION

The authors propose a novel approach for the segmentation of myocardium from DE MRI by using the sparse representation-based shape model and guided random walks. The sparse representation method effectively models the prior shape with a small number of shape templates, and the proposed method has the potential to achieve clinically relevant results.

Unsupervised CT Lung Image Segmentation of a Mycobacterium Tuberculosis Infection Model

Tuberculosis (TB) is an infectious disease caused by Mycobacterium tuberculosis that produces pulmonary damage. Radiological imaging is the preferred technique for the assessment of TB longitudinal course. Computer-assisted identification of biomarkers eases the work of the radiologist by providing a quantitative assessment of disease. Lung segmentation is the step before biomarker extraction. In this study, we present an automatic procedure that enables robust segmentation of damaged lungs that have lesions attached to the parenchyma and are affected by respiratory movement artifacts in a Mycobacterium Tuberculosis infection model. Its main steps are the extraction of the healthy lung tissue and the airway tree followed by elimination of the fuzzy boundaries. Its performance was compared with respect to a segmentation obtained using: (1) a semi-automatic tool and (2) an approach based on fuzzy connectedness. A consensus segmentation resulting from the majority voting of three experts' annotations was considered our ground truth. The proposed approach improves the overlap indicators (Dice similarity coefficient, 94% ± 4%) and the surface similarity coefficients (Hausdorff distance, 8.64 mm ± 7.36 mm) in the majority of the most difficult-to-segment slices. Results indicate that the refined lung segmentations generated could facilitate the extraction of meaningful quantitative data on disease burden.

Human Part Segmentation in Depth Images with Annotated Part Positions

We present a method of segmenting human parts in depth images, when provided the image positions of the body parts. The goal is to facilitate per-pixel labelling of large datasets of human images, which are used for training and testing algorithms for pose estimation and automatic segmentation. A common technique in image segmentation is to represent an image as a two-dimensional grid graph, with one node for each pixel and edges between neighbouring pixels. We introduce a graph with distinct layers of nodes to model occlusion of the body by the arms. Once the graph is constructed, the annotated part positions are used as seeds for a standard interactive segmentation algorithm. Our method is evaluated on two public datasets containing depth images of humans from a frontal view. It produces a mean per-class accuracy of 93.55% on the first dataset, compared to 87.91% (random forest and graph cuts) and 90.31% (random forest and Markov random field). It also achieves a per-class accuracy of 90.60% on the second dataset. Future work can experiment with various methods for creating the graph layers to accurately model occlusion.

A topo-graph model for indistinct target boundary definition from anatomical images

BACKGROUND AND OBJECTIVE

It can be challenging to delineate the target object in anatomical imaging when the object boundaries are difficult to discern due to the low contrast or overlapping intensity distributions from adjacent tissues.

METHODS

We propose a topo-graph model to address this issue. The first step is to extract a topographic representation that reflects multiple levels of topographic information in an input image. We then define two types of node connections - nesting branches (NBs) and geodesic edges (GEs). NBs connect nodes corresponding to initial topographic regions and GEs link the nodes at a detailed level. The weights for NBs are defined to measure the similarity of regional appearance, and weights for GEs are defined with geodesic and local constraints. NBs contribute to the separation of topographic regions and the GEs assist the delineation of uncertain boundaries. Final segmentation is achieved by calculating the relevance of the unlabeled nodes to the labels by the optimization of a graph-based energy function. We test our model on 47 low contrast CT studies of patients with non-small cell lung cancer (NSCLC), 10 contrast-enhanced CT liver cases and 50 breast and abdominal ultrasound images. The validation criteria are the Dice's similarity coefficient and the Hausdorff distance.

RESULTS

Student's t-test show that our model outperformed the graph models with pixel-only, pixel and regional, neighboring and radial connections (p-values <0.05).

CONCLUSIONS

Our findings show that the topographic representation and topo-graph model provides improved delineation and separation of objects from adjacent tissues compared to the tested models.

Radiomics in Nuclear Medicine Applied to Radiation Therapy: Methods, Pitfalls, and Challenges

Radiomics is a recent area of research in precision medicine and is based on the extraction of a large variety of features from medical images. In the field of radiation oncology, comprehensive image analysis is crucial to personalization of treatments. A better characterization of local heterogeneity and the shape of the tumor, depicting individual cancer aggressiveness, could guide dose planning and suggest volumes in which a higher dose is needed for better tumor control. In addition, noninvasive imaging features that could predict treatment outcome from baseline scans could help the radiation oncologist to determine the best treatment strategies and to stratify patients as at low risk or high risk of recurrence. Nuclear medicine molecular imaging reflects information regarding biological processes in the tumor thanks to a wide range of radiotracers. Many studies involving ¹⁸F-fluorodeoxyglucose positron emission tomography suggest an added value of radiomics compared with the use of conventional PET metrics such as standardized uptake value for both tumor diagnosis and prediction of recurrence or treatment outcome. However, these promising results should not hide technical difficulties that still currently prevent the approach from being widely studied or clinically used. These difficulties mostly pertain to the variability of the imaging features as a function of the acquisition device and protocol, the robustness of the models with respect to that variability, and the interpretation of the radiomic models. Addressing the impact of the variability in acquisition and reconstruction protocols is needed, as is harmonizing the radiomic feature calculation methods, to ensure the reproducibility of studies in a multicenter context and their implementation in a clinical workflow. In this review, we explain the potential impact of positron emission tomography radiomics for radiation therapy and underline the various aspects that need to be carefully addressed to make the most of this promising approach.

BRANE Clust: Cluster-Assisted Gene Regulatory Network Inference Refinement

Discovering meaningful gene interactions is crucial for the identification of novel regulatory processes in cells. Building accurately the related graphs remains challenging due to the large number of possible solutions from available data. Nonetheless, enforcing a priori on the graph structure, such as modularity, may reduce network indeterminacy issues. BRANE Clust (Biologically-Related A priori Network Enhancement with Clustering) refines gene regulatory network (GRN) inference thanks to cluster information. It works as a post-processing tool for inference methods (i.e., CLR, GENIE3). In BRANE Clust, the clustering is based on the inversion of a system of linear equations involving a graph-Laplacian matrix promoting a modular structure. Our approach is validated on DREAM4 and DREAM5 datasets with objective measures, showing significant comparative improvements. We provide additional insights on the discovery of novel regulatory or co-expressed links in the inferred Escherichia coli network evaluated using the STRING database. The comparative pertinence of clustering is discussed computationally (SIMoNe, WGCNA, X-means) and biologically (RegulonDB). BRANE Clust software is available at: http://www-syscom.univ-mlv.fr/~pirayre/Codes-GRN-BRANE-clust.html.

Latency Management in Scribble-Based Interactive Segmentation of Medical Images

OBJECTIVE

During an interactive image segmentation task, the outcome is strongly influenced by human factors. In particular, a reduction in computation time does not guarantee an improvement in the overall segmentation time. This paper characterizes user efficiency during scribble-based interactive segmentation as a function of computation time.

METHODS

We report a controlled experiment with users who experienced eight different levels of simulated latency (ranging from 100 to 2000 ms) with two techniques for refreshing visual feedback (either automatic, where the segmentation was recomputed and displayed continuously during label drawing, or user initiated, which was only computed and displayed each time the user hits a defined button).

RESULTS

For short latencies, the user's attention is focused on the automatic visual feedback, slowing down his/her labeling performance. This effect is attenuated as the latency grows larger, and the two refresh techniques yield similar user performance at the largest latencies. Moreover, during the segmentation task, participants spent in average for automatic refresh and for user-initiated refresh of the overall segmentation time interpreting the results.

CONCLUSION

The latency is perceived differently according to the refresh method used during the segmentation task. Therefore, it is possible to reduce its impact on the user performance.

SIGNIFICANCE

This is the first time a study investigates the effects of latency in an interactive segmentation task. The analysis and recommendations provided in this paper help understanding the cognitive mechanisms in interactive image segmentation.

Automated cell type discovery and classification through knowledge transfer

Motivation

Recent advances in mass cytometry allow simultaneous measurements of up to 50 markers at single-cell resolution. However, the high dimensionality of mass cytometry data introduces computational challenges for automated data analysis and hinders translation of new biological understanding into clinical applications. Previous studies have applied machine learning to facilitate processing of mass cytometry data. However, manual inspection is still inevitable and becoming the barrier to reliable large-scale analysis.

Results

We present a new algorithm called Automated Cell-type Discovery and Classification (ACDC) that fully automates the classification of canonical cell populations and highlights novel cell types in mass cytometry data. Evaluations on real-world data show ACDC provides accurate and reliable estimations compared to manual gating results. Additionally, ACDC automatically classifies previously ambiguous cell types to facilitate discovery. Our findings suggest that ACDC substantially improves both reliability and interpretability of results obtained from high-dimensional mass cytometry profiling data.

Availability and Implementation

A Python package (Python 3) and analysis scripts for reproducing the results are availability on https://bitbucket.org/dudleylab/acdc.

Supplementary information

Supplementary data are available at Bioinformatics online.

Appearance Constrained Semi-Automatic Segmentation from DCE-MRI is Reproducible and Feasible for Breast Cancer Radiomics: A Feasibility Study

We present a segmentation approach that combines GrowCut (GC) with cancer-specific multi-parametric Gaussian Mixture Model (GCGMM) to produce accurate and reproducible segmentations. We evaluated GCGMM using a retrospectively collected 75 invasive ductal carcinoma with ERPR+ HER2− (n = 15), triple negative (TN) (n = 9), and ER-HER2+ (n = 57) cancers with variable presentation (mass and non-mass enhancement) and background parenchymal enhancement (mild and marked). Expert delineated manual contours were used to assess the segmentation performance using Dice coefficient (DSC), mean surface distance (mSD), Hausdorff distance, and volume ratio (VR). GCGMM segmentations were significantly more accurate than GrowCut (GC) and fuzzy c-means clustering (FCM). GCGMM’s segmentations and the texture features computed from those segmentations were the most reproducible compared with manual delineations and other analyzed segmentation methods. Finally, random forest (RF) classifier trained with leave-one-out cross-validation using features extracted from GCGMM segmentation resulted in the best accuracy for ER-HER2+ vs. ERPR+/TN (GCGMM 0.95, expert 0.95, GC 0.90, FCM 0.92) and for ERPR + HER2− vs. TN (GCGMM 0.92, expert 0.91, GC 0.77, FCM 0.83).

Mean curvature and texture constrained composite weighted random walk algorithm for optic disc segmentation towards glaucoma screening

Accurate optic disc (OD) segmentation is an important step in obtaining cup-to-disc ratio-based glaucoma screening using fundus imaging. It is a challenging task because of the subtle OD boundary, blood vessel occlusion and intensity inhomogeneity. In this Letter, the authors propose an improved version of the random walk algorithm for OD segmentation to tackle such challenges. The algorithm incorporates the mean curvature and Gabor texture energy features to define the new composite weight function to compute the edge weights. Unlike the deformable model-based OD segmentation techniques, the proposed algorithm remains unaffected by curve initialisation and local energy minima problem. The effectiveness of the proposed method is verified with DRIVE, DIARETDB1, DRISHTI-GS and MESSIDOR database images using the performance measures such as mean absolute distance, overlapping ratio, dice coefficient, sensitivity, specificity and precision. The obtained OD segmentation results and quantitative performance measures show robustness and superiority of the proposed algorithm in handling the complex challenges in OD segmentation.