Accuracy evaluation of initialization-free registration for intraoperative 3D-navigation

Purpose An initialization-free approach for perioperative registration in functional endoscopic sinus surgery (FESS) is sought. The quality of surgical navigation relies on registration accuracy of preoperative images to the patient. Although landmark-based registration is fast, it is prone to human operator errors. This study evaluates the accuracy of two well-known methods for segmentation of the occipital bone from CT-images for use in surgical 3D-navigation. Method The occipital bone was segmented for registration without pre-defined correspondences, with the iterative closest point algorithm (ICP). The thresholding plus marching cubes segmentation (TMCS), and the deformable model segmentation (DMS) were compared quantitatively by overlaying the areas of the segmentations in cross-sectional slices, and visually by displaying the pointwise distances between the segmentations in a three-dimensional distance map relative to an expert manual segmentation, taken as a “ground truth”. Results Excellent correspondence between the two methods was achieved; the results showed, however, that the TMCS is closer to the “ground truth”. This is due to the sub-voxel accuracy of the marching cubes algorithm by definition, and the sensitivity of the DMS method to the choice of parameters. The DMS approach, as a gradient-based method, is insensitive to the thresholding initialization. For noisy images and soft tissue delineation a gradient-based method, like the deformable model, performs better. Both methods correspond within minute differences less than 4%. Conclusion These results will allow further minimization of human interaction in the planning phase for intraoperative 3D-navigation, by allowing to automatically create surface patches for registration purposes, ultimately allowing to build an initialization-free, fully automatic registration procedure for navigated Ear-, Nose-, Throat- (ENT) surgery.     

Prostate boundary detection and volume estimation using TRUS images for brachytherapy applications

Objective: Authors propose a semi-automatic segmentation algorithm for three-dimensional prostate boundary detection from trans-rectal ultrasound images. As a part of brachytherapy treatment with seeds for early stage prostate cancer, a patient’s prostate is scanned using a trans-rectal ultrasound probe, its boundary is manually outlined, and its volume is estimated for dosimetry purposes. Proposed algorithm requires a reduced amount of radiologist’s input, and thus speeds up the surgical procedure. Methods: The proposed segmentation algorithm utilizes texture differences between ultrasound images of the prostate and the surrounding tissues. It is carried out in the polar coordinate system and uses three-dimensional data correlation to improve the smoothness and reliability of the segmentation. The algorithm is applied to axial trans-rectal ultrasound images and the results are compared to the “ground truth” set by manual prostate boundary outlining (by experienced radiologist). Method is validated on six patients. Results: In our tests, the proposed algorithm estimated prostate volume within 95% of the original radiologist’s estimate. Conclusions: The boundary segmentation obtained from the algorithm can reduce manual input by a factor of 3, without significantly affecting the accuracy of the segmentation. The reduction in the manual input reduces the overall brachytherapy procedure time.     

A probabilistic framework for freehand 3D ultrasound reconstruction applied to catheter ablation guidance in the left atrium

Introduction  The catheter ablation procedure is a minimally invasive surgery used to treat atrial fibrillation. Difficulty visualizing the catheter inside the left atrium anatomy has led to lengthy procedure times and limited success rates. In this paper, we present a set of algorithms for reconstructing 3D ultrasound data of the left atrium in real-time, with an emphasis on automatic tissue classification for improved clarity surrounding regions of interest. Methods  Using an intracardiac echo (ICE) ultrasound catheter, we collect 2D-ICE images of a left atrium phantom from multiple configurations and iteratively compound the acquired data into a 3D-ICE volume. We introduce two new methods for compounding overlapping US data—occupancy-likelihood and response-grid compounding—which automatically classify voxels as “occupied” or “clear,” and mitigate reconstruction artifacts caused by signal dropout. Finally, we use the results of an ICE-to-CT registration algorithm to devise a response-likelihood weighting scheme, which assigns weights to US signals based on the likelihood that they correspond to tissue-reflections. Results  Our algorithms successfully reconstruct a 3D-ICE volume of the left atrium with voxels classified as “occupied” or “clear,” even within difficult-to-image regions like the pulmonary vein openings. We are robust to dropout artifact that plagues a subset of the 2D-ICE images, and our weighting scheme assists in filtering out spurious data attributed to ghost-signals from multi-path reflections. By automatically classifying tissue, our algorithm precludes the need for thresholding, a process that is difficult to automate without subjective input. Our hope is to use this result towards developing 3D ultrasound segmentation algorithms in the future.     

基于MeanShift方法的肝脏CT图像的自动分割

目的 探讨基于Mean Shift方法的肝脏CT图像的自动分割算法,以实现肝脏的自动分割。方法 首先对原始图像进行单次Mean Shift平滑 ,滤除噪声的影响以增强算法的鲁棒性,然后通过Mean Shift迭代自动选取初始种子点,最后采用基于区域生长的方法实现肝脏CT图像的自动分割。结果 实验证明此方法是一个准确、快速和有效的肝脏自动分割方法。结论 采用本文中提出的方法,可有效地实现肝脏的自动分割。     

Joint level-set shape modeling and appearance modeling for brain structure segmentation

This paper presents a new fully automatic model-based segmentation algorithm, which combines level-set methods to model the shape of brain structures and their variation with active appearance modeling to generate images that are used to drive the segmentation. The new algorithm incorporates multi-modality images to improve the segmentation performance and the recursive least square (RLS) algorithm is adopted to minimize the difference between test image and the one synthesized from the shape and appearance modeling. When compared with manual segmentation, the 2D and 3D experiments demonstrate that the new algorithm is computationally efficient and robust and is promising for automatic segmentation of the lateral ventricles.     

Multi-task learning for the segmentation of organs at risk with label dependence

Automatic segmentation of organs at risk is crucial to aid diagnoses and remains a challenging task in medical image analysis domain. To perform the segmentation, we use multi-task learning (MTL) to accurately determine the contour of organs at risk in CT images. We train an encoder-decoder network for two tasks in parallel. The main task is the segmentation of organs, entailing a pixel-level classification in the CT images, and the auxiliary task is the multi-label classification of organs, entailing an image-level multi-label classification of the CT images. To boost the performance of the multi-label classification, we propose a weighted mean cross entropy loss function for the network training, where the weights are the global conditional probability between two organs. Based on MTL, we optimize the false positive filtering (FPF) algorithm to decrease the number of falsely segmented organ pixels in the CT images. Specifically, we propose a dynamic threshold selection (DTS) strategy to prevent true positive rates from decreasing when using the FPF algorithm. We validate these methods on the public ISBI 2019 segmentation of thoracic organs at risk (SegTHOR) challenge dataset and a private medical organ dataset. The experimental results show that networks using our proposed methods outperform basic encoder-decoder networks without increasing the training time complexity.     

Partial ring artifact on cardiac CT: image presentation and clinical implication

In computed tomography (CT), the term “artifact” is applied to any systematic discrepancy between the CT numbers in the reconstructed image and the true attenuation coefficients of the object. A partial ring artifact in cardiac CT has a unique pattern and has not been previously reported in real clinical practice. In this report, we will demonstrate the images and animations of a cardiac CT which is affected by a partial ring artifact due to a broken detector module. The different characteristics of ring artifacts on routine and cardiac CT, and the implications for clinical interpretation will also be described. In summary, when a partial ring artifact is noted incidentally on cardiac CT, we suggest interpreting the images by reviewing reconstructed images from different RR-intervals to avoid unnecessary repeated scans and additional radiation exposure.     

Peripancreatic lymphatic invasion by pancreatic carcinoma: evaluation with multi-detector row CT

Peripancreatic lymphatic networks are frequently involved in pancreatobiliary carcinoma, affecting the prognosis. However, little attention has been paid to CT imaging of normal and pathological conditions of peripancreatic lymphatic networks. We evaluated multi-detector row CT (MDCT) images of peripancreatic lymphatic networks invaded by pancreatic carcinoma and compared them with those of normal peripancreatic lymphatic networks using imaging reconstruction every 1 mm with a multiplanar reformation technique. Apart from the region around the pancreatic body and tail, normal peripancreatic lymphatic networks were detected as “linear structures” on MDCT. However, peripancreatic lymphatic invasion by pancreatic carcinoma was frequently identified as “reticular,” “tubular,” or “soft tissue mass” appearances in the peripancreatic fat tissues. Peripancreatic lymphatic invasion by pancreatic carcinoma was more frequently detected around the common hepatic artery, celiac artery, superior mesenteric artery, and left para-aortic area. Depending on the tumor location, positive peripancreatic lymphatic invasion was most frequent at the area around the common hepatic artery in the head region and at the area around the celiac artery in the body and tail regions. Knowledge of CT imaging of normal and pathological peripancreatic lymphatic networks is essential for determining the accurate staging of pancreatic carcinoma.     

Aorto-enteric fistula: CT findings

The objective of this pictorial essay is to present the different CT findings encountered in patients with aorto-enteric fistulas (AEFs). An AEF is a rare and sometimes disastrous occurrence, responsible for intermittent or massive gastrointestinal hemorrhage and hematemesis. CT provides fast and effective evaluation in hemodynamically stable patients suspected of having an AEF. The work was done in the CT department of “Korgialenio Benakio” Red Cross Hospital of Athens (Athanasaki 1 St., 11526 Athens, Greece).     

Biliary ascariasis: CT, MR cholangiopancreatography, and navigator endoscopic appearance—report of a case of acute biliary obstruction

A case of acute common bile duct obstruction caused by a roundworm in a 6-year-old girl is reported. Computed tomography (CT) and magnetic resonance imaging (MRI) showed “bull's-eye” and “eye-glass” appearances of the ascaris in the common bile duct. On reformation of the transverse CT and coronal MR images, the tubular Ascaris was better depicted. MRI and navigator endoscopic demonstrations of the common bile duct ascaris have not been described previously in the literature. These CT and MRI findings may be helpful in the diagnosis of Ascaris. Received 6 November 1998/Accepted 14 December 1998     

Automated histogram-based brain segmentation in T1-weighted three-dimensional magnetic resonance head images

Current semiautomated magnetic resonance (MR)-based brain segmentation and volume measurement methods are complex and not sufficiently accurate for certain applications. We have developed a simpler, more accurate automated algorithm for whole-brain segmentation and volume measurement in T(1)-weighted, three-dimensional MR images. This histogram-based brain segmentation (HBRS) algorithm is based on histograms and simple morphological operations. The algorithm's three steps are foreground/background thresholding, disconnection of brain from skull, and removal of residue fragments (sinus, cerebrospinal fluid, dura, and marrow). Brain volume was measured by counting the number of brain voxels. Accuracy was determined by applying HBRS to both simulated and real MR data. Comparing the brain volume rendered by HBRS with the volume on which the simulation is based, the average error was 1.38%. By applying HBRS to 20 normal MR data sets downloaded from the Internet Brain Segmentation Repository and comparing them with expert segmented data, the average Jaccard similarity was 0.963 and the kappa index was 0.981. The reproducibility of brain volume measurements was assessed by comparing data from two sessions (four total data sets) with human volunteers. Intrasession variability of brain volumes for sessions 1 and 2 was 0.55 +/- 0.56 and 0.74 +/- 0.56%, respectively; the mean difference between the two sessions was 0.60 +/- 0.46%. These results show that the HBRS algorithm is a simple, fast, and accurate method to determine brain volume with high reproducibility. This algorithm may be applied to various research and clinical investigations in which brain segmentation and volume measurement involving MRI data are needed.     

Automated aorta segmentation in low-dose chest CT images

Purpose

   Abnormalities of aortic surface and aortic diameter can be related to cardiovascular disease and aortic aneurysm. Computer-based aortic segmentation and measurement may aid physicians in related disease diagnosis. This paper presents a fully automated algorithm for aorta segmentation in low-dose non-contrast CT images.

Methods

   The original non-contrast CT scan images as well as their pre-computed anatomy label maps are used to locate the aorta and identify its surface. First a seed point is located inside the aortic lumen. Then, a cylindrical model is progressively fitted to the 3D image space to track the aorta centerline. Finally, the aortic surface is located based on image intensity information. This algorithm has been trained and tested on 359 low-dose non-contrast CT images from VIA-ELCAP and LIDC public image databases. Twenty images were used for training to obtain the optimal set of parameters, while the remaining images were used for testing. The segmentation result has been evaluated both qualitatively and quantitatively. Sixty representative testing images were used to establish a partial ground truth by manual marking on several axial image slices.

Results

   Compared to ground truth marking, the segmentation result had a mean Dice Similarity Coefficient of 0.933 (maximum 0.963 and minimum 0.907). The average boundary distance between manual segmentation and automatic segmentation was 1.39 mm with a maximum of 1.79 mm and a minimum of 0.83 mm.

Conclusion

   Both qualitative and quantitative evaluations have shown that the presented algorithm is able to accurately segment the aorta in low-dose non-contrast CT images.     

CRUISE: cortical reconstruction using implicit surface evolution

Segmentation and representation of the human cerebral cortex from magnetic resonance (MR) images play an important role in neuroscience and medicine. A successful segmentation method must be robust to various imaging artifacts and produce anatomically meaningful and consistent cortical representations. A method for the automatic reconstruction of the inner, central, and outer surfaces of the cerebral cortex from T1-weighted MR brain images is presented. The method combines a fuzzy tissue classification method, an efficient topology correction algorithm, and a topology-preserving geometric deformable surface model (TGDM). The algorithm is fast and numerically stable, and yields accurate brain surface reconstructions that are guaranteed to be topologically correct and free from self-intersections. Validation results on real MR data are presented to demonstrate the performance of the method.     

Coronary Artery CT Angiography for Plaque Imaging

Cardiac CT is becoming a mainstream and integral part of many cardiology practices based on a vast base of literature supporting and validating its clinical utility. As the technology continues to advance, coronary imaging has improved in stride. In the next several years, cardiac CT may become the “gatekeeper” of cardiac testing, surpassing the more common and widespread nuclear testing as the initial strategy in evaluating ischemia. Unfortunately, in spite of an arsenal of tests available to detect clinically significant stable coronary artery disease, many people continue to suffer acute myocardial infarction and other acute coronary syndromes, leading to significant morbidity and mortality due to unstable coronary artery disease. These unstable, “vulnerable” plaques continue to plague cardiologists across the globe. The ability to identify vulnerable plaque is a step in the right direction toward therapy. It is in this particular arena that advancements in cardiac CT technology may bear the most fruit. A growing body of evidence supporting the utility of cardiac CT in plaque imaging has emerged and has demonstrated that potentially unstable coronary artery disease is able to be identified accurately and noninvasively.     

Neuromuscular electrical stimulation for a patient with metastatic lung cancer—a case report

A 47-year-old female patient suffering from advanced lung cancer with metastatic bone and brain disease participated in a passive exercise program, consisting of neuromuscular electrical stimulation (NMES) five times a week, carried out for 4 weeks. After the training period, the results of the 6-min walk (420 m before and 603 m after the training period) have improved by 44%, which demonstrates the increase of physical performance (mobility, endurance capacity). The results of the “Timed up and go” indicate an improvement of mobility and functional health of skeletal muscles. Furthermore, the quality of life (QOL)-scales (assessed by using the SF-36 health survey) “Physical functioning”, “Role—physical”, “Mental health”, “Role—emotional”, “Vitality”, “Bodily pain”, and “General health” showed improvements after the intervention period. Feasibility, safety, and beneficial effects of the NMES program were proven for the patient in this case study. These findings indicate that NMES, initiated and executed with appropriate care, may serve as a useful supportive means of palliative treatment in some patients with advanced cancer and metastatic disease, especially in cases of metastatic involvement of the brain and of the skeletal system with the risk of seizures and pathological fractures where volitional training is not allowed.     

Schistosomiasis of the liver

Schistosomiasis is an infection of trematodes, Schistosoma, causing periportal fibrosis and liver cirrhosis due to deposition of eggs in the small portal venules. In schistosomiasis caused by S. mansoni, sonography shows echogenic thickening or fibrotic band along the portal veins. CT shows low-attenuation bands or rings around the large portal vein branches in the central part of the liver with marked enhancement. Hepatoplenomegaly, liver cirrhosis, portal hypertension and gastroesophageal varies are commonly associated. In schistosomiasis caused by S. japonicum, sonography shows echogenic septae in the liver, utlining the polygonal liver lobules, mimicking “fish-scale” network appearance, reflecting fibrosis. CT shows periportal septae in the peripheral part of the liver parenchyma, producing “turtle-back” appearance, representing calcified eggs along the portal tracts. The portal tracts and hepatic capsule are enhanced on contrast-enhanced CT images. The size and shape of the liver are relatively preserved. MR images show fibrous septae as low signal intensity on T1-weighted images, high signal intensity on T2-weighted images, and these fibrous septae are enhanced. CT images of the lungs show multiple scattered nodules with halo of ground-glass opacities. Exudative granulomatous inflammation of the colonic wall may produce inflammatory polyps, fibrous thickening or stenosis of the colonic wall.     

Low-rank and sparse decomposition based shape model and probabilistic atlas for automatic pathological organ segmentation

One major limiting factor that prevents the accurate delineation of human organs has been the presence of severe pathology and pathology affecting organ borders. Overcoming these limitations is exactly what we are concerned in this study. We propose an automatic method for accurate and robust pathological organ segmentation from CT images. The method is grounded in the active shape model (ASM) framework. It leverages techniques from low-rank and sparse decomposition (LRSD) theory to robustly recover a subspace from grossly corrupted data. We first present a population-specific LRSD-based shape prior model, called LRSD-SM, to handle non-Gaussian gross errors caused by weak and misleading appearance cues of large lesions, complex shape variations, and poor adaptation to the finer local details in a unified framework. For the shape model initialization, we introduce a method based on patient-specific LRSD-based probabilistic atlas (PA), called LRSD-PA, to deal with large errors in atlas-to-target registration and low likelihood of the target organ. Furthermore, to make our segmentation framework more efficient and robust against local minima, we develop a hierarchical ASM search strategy. Our method is tested on the SLIVER07 database for liver segmentation competition, and ranks 3rd in all the published state-of-the-art automatic methods. Our method is also evaluated on some pathological organs (pathological liver and right lung) from 95 clinical CT scans and its results are compared with the three closely related methods. The applicability of the proposed method to segmentation of the various pathological organs (including some highly severe cases) is demonstrated with good results on both quantitative and qualitative experimentation; our segmentation algorithm can delineate organ boundaries that reach a level of accuracy comparable with those of human raters.     

Skull-stripping magnetic resonance brain images using a model-based level set

The segmentation of brain tissue from nonbrain tissue in magnetic resonance (MR) images, commonly referred to as skull stripping, is an important image processing step in many neuroimage studies. A new mathematical algorithm, a model-based level set (MLS), was developed for controlling the evolution of the zero level curve that is implicitly embedded in the level set function. The evolution of the curve was controlled using two terms in the level set equation, whose values represented the forces that determined the speed of the evolving curve. The first force was derived from the mean curvature of the curve, and the second was designed to model the intensity characteristics of the cortex in MR images. The combination of these forces in a level set framework pushed or pulled the curve toward the brain surface. Quantitative evaluation of the MLS algorithm was performed by comparing the results of the MLS algorithm to those obtained using expert segmentation in 29 sets of pediatric brain MR images and 20 sets of young adult MR images. Another 48 sets of elderly adult MR images were used for qualitatively evaluating the algorithm. The MLS algorithm was also compared to two existing methods, the brain extraction tool (BET) and the brain surface extractor (BSE), using the data from the Internet brain segmentation repository (IBSR). The MLS algorithm provides robust skull-stripping results, making it a promising tool for use in large, multi-institutional, population-based neuroimaging studies.     

Multi-scale patch and multi-modality atlases for whole heart segmentation of MRI

A whole heart segmentation (WHS) method is presented for cardiac MRI. This segmentation method employs multi-modality atlases from MRI and CT and adopts a new label fusion algorithm which is based on the proposed multi-scale patch (MSP) strategy and a new global atlas ranking scheme. MSP, developed from the scale-space theory, uses the information of multi-scale images and provides different levels of the structural information of images for multi-level local atlas ranking. Both the local and global atlas ranking steps use the information theoretic measures to compute the similarity between the target image and the atlases from multiple modalities. The proposed segmentation scheme was evaluated on a set of data involving 20 cardiac MRI and 20 CT images. Our proposed algorithm demonstrated a promising performance, yielding a mean WHS Dice score of 0.899 ± 0.0340, Jaccard index of 0.818 ± 0.0549, and surface distance error of 1.09 ± 1.11 mm for the 20 MRI data. The average runtime for the proposed label fusion was 12.58 min.     

The choledochal ring sign: a specific finding in acute biliary pancreatitis

We investigated a marked common bile duct wall contrast-enhancement in acute pancreatitis on CT scans as an accurate sign of biliary pancreatitis. Contrast-enhanced CT scans of 80 patients with clinically and biologically confirmed acute pancreatitis were reviewed by two gastrointestinal radiologists without knowledge of the cause of acute pancreatitis. Since this study was retrospective and then did not modify the CT examination routinely performed in acute abdomen, no institutional review board approval and informed consent was requested. A systematic measure of the difference between the CT number of common bile duct wall and the CT number of pancreatic parenchyma was performed. The “choledochal ring” sign, defined by a difference greater than 15 HU was specifically studied and compared to other factors often associated with biliary pancreatitis through univariate and multivariate analyses. Compared to other factors classically associated with acute biliary pancreatitis, the “choledochal ring” sign showed high specificity (100%) and high positive predictive value (100%). In patients with acute pancreatitis, the “choledochal ring” sign is a new and accurate finding of acute biliary pancreatitis on CT scans.