Contour Of Region Research Articles

How can we measure the renal pelvic anteroposterior diameter in postnatal isolated hydronephrosis?

The anteroposterior diameter (APD) of the renal pelvis is a commonly used ultrasound parameter in the evaluation and management of hydronephrosis. It has been established that an APD value associated with pyeloplasty is around 25mm. Some believe the APD should be measured at the innermost part of the renal pelvis while others suggest that it should be done at the renal contour. However, there is no consensus on the optimal APD measurement technique including whether it should be measured supine or prone. This study compared six different techniques of APD measurements, in both supine and prone positions, and further evaluated their association with pyeloplasty. Data was obtained by retrospectively reviewing patients' charts that had initial high-grade hydronephrosis (HGH) from 2008 to 2014. We recorded the patients' demographics, ultrasound data and management choice. In the mid-renal transverse plane, the APD was measured by 2 blinded investigators, at the intra-renal, renal contour and extra-renal regions of the renal pelvis in supine and prone positions (FigureA). We compared the six APD measurements based on the outcome of management (pyeloplasty vs. conservative management). The ROC curve obtained was then used to assess the ability of various APD measurements in predicting surgical intervention. The cutoff value chosen that predicts pyeloplasty was the lowest diameter with 100% specificity. We included 129 patients (134 renal units). Forty-four renal units (42 patients) underwent pyeloplasty whereas 90 renal units (87 patients) were managed conservatively. Patients' demographics were grouped by both SFU grade and clinical outcome. Regardless of grade, the APD measurements were different in all 6 techniques. All APD measurement techniques showed good inter-rater reliability. Based on the ROC curve, all APD measurements were associated with pyeloplasty with an AUC from 0.89 to 0.91. The supine extra-renal APD measurement of 24mm was the most sensitive cutoff value. The cutoff values ranged from 18 to 27mm when including patients from all grades of hydronephrosis. The median APD measurements were significantly less for SFU grade 3 than grade 4 hydronephrosis in all positions (P<0.001 for all measurements), yet the predictive cutoff value of 24mm for the supine extra-renal was similar for both grades. APD measurements differ based on the technique, but they are all equally associated with the clinical outcome of pyeloplasty. The inter-rater reliability of all techniques were excellent. Though the median APD measurements are smaller in SFU grade 3, it appears that the cutoff for a predictive renal pelvic APD does not differ between SFU grades 3 and 4 for the supine extra-renal technique. We conclude that the technique for measuring the APD needs to be specified in studies of hydronephrosis and in any grading systems.

Confidence Region Identification and Contour Detection in MRI Image.

Tumour region extraction (RE) method identifies the area of interest in MR imaging as it also highlights tumour boundaries. Some other intensities are existing, they are not visible but have their existence in region, and this region is called growing region. Such region is to be tumour region. Due to the variation of intensities in MRI images, tumour visibility becomes uncleared. Tumour intensity variations (tumour tissues) mix with normal brain tissues. In the light of above circumstance, tumour growing region becomes challenge. The goal of work is to extract the region of interest with confidence. The objective of the study is to develop the region of interest of brain tumour MRI image method by using confidence score for identifying the variation of intensity. The significance of work is based on identification of region of interest (tumour region). Confidence score is measured through pattern of intensities of MRI image. Similar patterns of brain tumour intensities are identified. Each pattern of intensities is adjusted with certain scale, and then biggest blob is analysed. Various biggest area blobs are combined, and resultant biggest blob is formed. In fact, resultant area blob is a combination of different patterns. Each pattern is assigned with particular colour. These colours highlight the growing region. Further, a contour is detected around the tumour boundaries. With combination of region scale fitting and contour detection (CD), tumour boundaries are further separated from normal tissues. Hence, the confidence score (CS) is formed from CD. CS is further converted to confidence region (CR). Conversion to CR is performed though confidence interval (CI). CI is based on defined conditions. In such conditions, different probabilities are considered. Probability identifies the region. Source of region formation is pixels; these pixels highlight tumour core significantly. This CR is obtained through checking standard deviation and statistical evaluation using confidence interval. Hence, region-of-interest pixels are identifying the CR. CR is evaluated through 97% Dice over index (DOI), 94% Jacquard, MSE 1.24, and PSNR 17.45. Value of testing parameter from benchmark study was JI, DOI, and MSE, PSNR : JI was 31.5%, DOI was 47.3%, MSE was 2.5 dB, and PSNR was 40 dB. The parameters are measured for the complex images; contribution parameter classifies the mean pixel values and deviating pixel values, and the classification of the pixel value is like to be termed as intensities. Mentioned classification extracts the variation of intensity pixels accurately; then, algorithm is highlighting the region as compared to the normal tumour cells.

Semantic Segmentation for Free Space and Lane Based on Grid-Based Interest Point Detection

An increasing number of tasks have been developed for autonomous driving and advanced driver assistance systems. However, this gives rise to the problem of incorporating plural functionalities to be ported into a power-constrained computing device. Therefore, the objective of this work is to alleviate the complex learning procedure of the pixel-wise approach for driving scene understanding. In this paper, we go beyond the pixel-wise detection of the semantic segmentation task as a point detection task and implement it to detect free space and lane. Instead of pixel-wise learning, we trained a single deep convolution neural network for point of interest detection in a grid-based level and followed with a computer vision (CV) based post-processing of end branches corresponding to the characteristic of target classes. To achieve the corresponding final result of pixel-wise detection of semantic segmentation and parametric description of lanes, we propose a CV-based post-processing to decode points of output from the neural network. The final results showed that the network could learn the spatial relationship for point of interest, including the representative points on the contour of the free space segmented region and the representative points along the center of the road lane. We verify our method on two publicly available datasets, which achieved 98.2% mIoU on the KITTI dataset for the evaluation of free space and 97.8% accuracy on the TuSimple dataset (with the field of view below the <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$y=320$ </tex-math></inline-formula> axis) for the evaluation of the lane.

Методы формирования ортогональных многогранников для задач раскроя и упаковки объектов произвольной геометрии

The article deals with the problem of packing objects of arbitrary geometry. Modern methods of designing irregular packing schemes use a mathematical model based on phi-functions and a hodograph vector function of dense placement. These methods make it possible to obtain exact solutions, but they are time-consuming and very sensitive to the dimension of the problem being solved and the degree of detail of the geometry of vector objects. The use of a discrete representation of placed objects in the form of orthogonal polyhedra can signifi increase the speed of construction a packing, which makes the problem of adequately transforming the shape of placed objects (vector models in the two-dimensional case and polygonal models in the three- dimensional case) relevant. The aim of the study is to systematize methods that provide the formation of orthogonal polyhedra of various dimensions for describing objects and containers of arbitrary geometry. Methods for creating orthogonal polyhedra based on set-theoretic operations (addition, subtraction and intersection), analytical modeling using a set of functions and relational operators, as well as voxelization of fl and volumetric object models are considered. The use of set-theoretic operations is best suited for the manual creation of orthogonal polyhedra with relatively simple geometry. The method of analytical modeling is intended for the formation of voxelized objects based on geometric fi es described by a set of analytically specifi functions. The application of various relational operators to obtain orthogonal polyhedra that describe the contour, internal and external regions of analytical given objects is shown. An algorithm for creating a container in the form of an orthogonal polyhedron based on a given vector model is proposed, which makes it possible to solve problems of irregular packing of objects inside containers of arbitrary shape. All the methods presented in the article are programmatically implemented with a generalization in terms of dimension and are applicable to solving any types of cutting and packing problems.

Scene-graph-driven semantic feature matching for monocular digestive endoscopy

Background and objectiveRegistration of the preoperative 3D model with the video of the digestive tract is the key task in endoscopy surgical navigation. Accurate 3D reconstruction of soft tissue surfaces is essential to complete registration. However, existing feature matching methods still fall short of desirable performance, due to the soft tissue deformation and smooth but less-textured surface. MethodsIn this paper, we present a new semantic description based on the scene graph to integrate contour features and SIFT features. Firstly, we construct the semantic feature descriptor using the SIFT features and dense points in the contour regions to obtain more dense point feature matching. Secondly, we design a clustering algorithm based on the proposed semantic feature descriptor. Finally, we apply the semantic description to the structure from motion (SfM) reconstruction framework. ResultsOur techniques are validated by the phantom tests and real surgery videos. We compare our approaches with other typical methods in contour extraction, feature matching, and SfM reconstruction. On average, the feature matching accuracy reaches 75.6% and improves 16.6% in pose estimation. In addition, 39.8% of sparse points are increased in SfM results, and 35.31% more valid points are obtained for the DenseDescriptorNet training in 3D reconstruction. ConclusionsThe new semantic feature description has the potential to reveal more accurate and dense feature correspondence and provides local semantic information in feature matching. Our experiments on the clinical dataset demonstrate the effectiveness and robustness of the novel approach.

Priors-guided convolutional neural network for 3D foveal avascular zone segmentation.

The foveal avascular zone (FAZ) is sensitive to retinal pathological process in the macular fovea area. For the purpose of efficient FAZ 3D quantification, we firstly propose a priors-guided convolutional neural network (CNN) to provide a tailor-made solution for 3D FAZ segmentation for optical coherence tomography angiography (OCTA) images. Location and topology priors are taken into account. The random central crop module is utilized to restrict the region to be processed, while the non-local attention gates are contained in the network to capture long-range dependency. The topological consistency constraint is calculated on maximum and mean projection maps through persistent homology to keep topological correctness of the model's prediction. Our method was evaluated on two OCTA datasets with 478 eyes and the experimental results demonstrate that our method can not only alleviate the over-segmentation prominently but also fit better on the contour of FAZ region.

Magnetic resonance imaging contrast enhancement synthesis using cascade networks with local supervision.

Gadolinium-based contrast agents (GBCAs) are widely administrated in MR imaging for diagnostic studies and treatment planning. Although GBCAs are generally thought to be safe, various health and environmental concerns have been raised recently about their use in MR imaging. The purpose of this work is to derive synthetic contrast enhance MR images from unenhanced counterpart images, thereby eliminating the need for GBCAs, using a cascade deep learning workflow that incorporates contour information into the network. The proposed workflow consists of two sequential networks: (1) a retina U-Net, which is first trained to derive semantic features from the non-contrast MR images in representing the tumor regions; and (2) a synthesis module, which is trained after the retina U-Net to take the concatenation of the semantic feature maps and non-contrast MR image as input and to generate the synthetic contrast enhanced MR images. After network training, only the non-contrast enhanced MR images are required for the input in the proposed workflow. The MR images of 369 patients from the multimodal brain tumor segmentation challenge 2020 (BraTS2020) dataset were used in this study to evaluate the proposed workflow for synthesizing contrast enhanced MR images (200 patients for five-fold cross-validation and 169 patients for hold-out test). Quantitative evaluations were conducted by calculating the normalized mean absolute error (NMAE), structural similarity index measurement (SSIM), and Pearson correlation coefficient (PCC). The original contrast enhanced MR images were considered as the ground truth in this analysis. The proposed cascade deep learning workflow synthesized contrast enhanced MR images that are not visually differentiable from the ground truth with and without supervision of the tumor contours during the network training. Difference images and profiles of the synthetic contrast enhanced MR images revealed that intensity differences could be observed in the tumor region if the contour information was not incorporated in network training. Among the hold-out test patients, mean values and standard deviations of the NMAE, SSIM, and PCC were 0.063±0.022, 0.991±0.007 and 0.995±0.006, respectively, for the whole brain; and were 0.050±0.025, 0.993±0.008 and 0.999±0.003, respectively, for the tumor contour regions. Quantitative evaluations with five-fold cross-validation and hold-out test showed that the calculated metrics can be significantly enhanced (p-values ≤ 0.002) with the tumor contour supervision in network training. The proposed workflow was able to generate synthetic contrast enhanced MR images that closely resemble the ground truth images from non-contrast enhanced MR images when the network training included tumor contours. These results suggest that it may be possible to minimize the use of GBCAs in cranial MR imaging studies.

Dim and Small Target Detection Based on Improved Spatio-Temporal Filtering

Small target detection in high strength clutter background is in great in remote imaging system, a new improved spatio-temporal filtering was proposed in this paper. Firstly, traditional anisotropy filtering has poor suppression effect in strength edge contour region, so a new diffusion filtering function proposed in paper. According to the analysis with difference of each component of the image, a new anisotropy diffusion function is constructed in this paper. When the difference of background and target is small, this algorithm will give in large diffusion coefficient to filter most background clutter and retain target signal well which achieves background prediction better. Secondly, because the traditional spatiotemporal filter algorithm cant follow the motion object in the fixed search pipe diameter what will make lose the target detection, a new weight constraint function of adaptive change of the search diameter in this paper is built which can change the search diameter with the moving of target, and improve the detection accuracy. Finally, experiments show that compared with traditional algorithms and detected in different scenes, this method can enhance small target detection effectively.

Multi-dimension tool wear state assessment criterion on the spiral edge of the milling cutter

In the assessment of the tool wear state for the spiral edge of milling cutter based on machine vision, the traditional assessment criterion is often inaccurate due to the problem of missing of information, especially near the tip area. In order to deal with this problem, different lighting-condition settings and additional non-image information compensation techniques are needed. In view of this, an integration tool wear detection method combined line laser edge detection and machine vision is proposed. Subsequently, the combined data acquirement experimental system is designed and built, which can simultaneously acquire two types of parameters: diameter and images under the same detection condition. Then, based on the above information, a multi-dimension series assessment criterion is proposed consisting of three dimensions index: The one-dimensional assessment index gives the average wear value of the spiral side, which is mainly used as data ordination and characterizes the rule of tool wear degradation; the two-dimensional assessment index describes the contour of the wear region and calculates the area value, which can predict the change of the tool wear stage more precisely; and the three-dimensional assessment index gives the 3D morphology by adding depth information within the wear region and quantifies the volume of the worn-off part, which provides clues for early warning of the deterioration caused by the tool worn. Experiments proved that the multi-dimension series assessment criterion is helpful in reflecting the trend of tool life and giving a more accurate assessment of the tool wear state for the spiral edge of a milling cutter.

CRISP: a deep learning architecture for GC × GC-TOFMS contour ROI identification, simulation and analysis in imaging metabolomics.

Two-dimensional gas chromatography–time-of-flight mass spectrometry (GC × GC–TOFMS) provides a large amount of molecular information from biological samples. However, the lack of a comprehensive compound library or customizable bioinformatics tool is currently a challenge in GC × GC–TOFMS data analysis. We present an open-source deep learning (DL) software called contour regions of interest (ROI) identification, simulation and untargeted metabolomics profiler (CRISP). CRISP integrates multiple customizable deep neural network architectures for assisting the semi-automated identification of ROIs, contour synthesis, resolution enhancement and classification of GC × GC–TOFMS-based contour images. The approach includes the novel aggregate feature representative contour (AFRC) construction and stacked ROIs. This generates an unbiased contour image dataset that enhances the contrasting characteristics between different test groups and can be suitable for small sample sizes. The utility of the generative models and the accuracy and efficacy of the platform were demonstrated using a dataset of GC × GC–TOFMS contour images from patients with late-stage diabetic nephropathy and healthy control groups. CRISP successfully constructed AFRC images and identified over five ROIs to create a deepstacked dataset. The high fidelity, 512 × 512-pixels generative model was trained as a generator with a Fréchet inception distance of <47.00. The trained classifier achieved an AUROC of >0.96 and a classification accuracy of >95.00% for datasets with and without column bleed. Overall, CRISP demonstrates good potential as a DL-based approach for the rapid analysis of 4-D GC × GC–TOFMS untargeted metabolite profiles by directly implementing contour images. CRISP is available at https://github.com/vivekmathema/GCxGC-CRISP.

Multitarget Pedestrian Tracking Algorithm Based on a Contour Template with Multiscale Stability

In order to overcome the problems of high error rate and poor tracking effect of traditional tracking algorithms, a multitarget pedestrian tracking algorithm based on a contour template is designed in this paper. The pedestrian template is divided into several contour regions by template voting strategy and pedestrian features are classified by regional feature similarity classification. Then, the score function is selected to realize the multiobjective pedestrian feature division, and the multicontour template matching is realized according to the similarity of deformation diversity. According to the matching results, the detection and tracking of multitarget objects are realized. The experimental results show that the feature recognition accuracy of the algorithm is between 93% and 98%, the tracking error of pedestrian gravity center position is always below 5%, and the tracking time is always below 5 S in different video frame scales, which fully proves the effectiveness of the algorithm.

A Comprehensive Approach To Alloplastic Cranioplasty: Novel Implant Design And Refinement of Soft Tissue Management

Background: Large scale craniectomy defects are commonly reconstructed with alloplastic implants, which can restore brain protection and promote cosmesis. However, esthetic outcomes can be subpar due to skin contour abnormalities and temporal hollowing. Herein we describe a senior craniofacial surgeon’s experience using a custom composite polyetheretherketone (PEEK) and porous polyethylene (Medpor) implant for alloplastic cranioplasty. Methods: A retrospective review was conducted of all PEEK-Medpor cranioplasty cases performed over the past 2 years. Patient characteristics, intraoperative information, surgical outcomes, and cosmetic outcomes were reviewed. Results: Sixteen patients (18-70 years of age) underwent surgery. Indications for cranioplasty included craniectomy due to trauma, stroke, or tumor, and bone resorption after a prior autologous cranioplasty. Augmentation of the temporalis area using alloderm or a muscle graft was performed in 7 cases. There were no intraoperative complications. Ultimately, all implants were maintained. One implant was temporarily removed due to infection, but successfully replaced. Three minor complications occurred. At an average follow-up of 9 months, 93.8% of surgical sites showed no significant temporal hollow. Conclusion: We describe a series of 16 implants using a composite PEEK Medpor implant for alloplastic cranioplasty, which resulted in a low infection rate and improved postoperative regional contour. Use of this implant with suspension of the temporalis muscle is a new technique that may allow for better adherence of the temporalis muscle to its anatomic position, while still providing good brain protection.

Bleeding contour detection for craniotomy

ObjectiveBleeding impairs observation during neurosurgery, and excessive bleeding endangers the life of a patient. Thus, hemostasis is important during neurosurgery. The detection of bleeding areas is a prerequisite for hemostasis. MethodsTo the best of our knowledge, this paper is the first to present results on the detection of neurosurgical craniotomy bleeding scenarios, i.e., scalp incision bleeding, skull incision bleeding, and dura matter-incision bleeding. This is realized via a workflow that combines craniotomy image data preparation and a Mask R-CNN framework. Bleeding images on a porcine skin tissue with a simulated blood injected by a syringe are taken by a visible light camera, and the video frames of the scalp incision, skull incision, and dura matter-incision bleeding are extracted from neurosurgical videos. ResultsThe precision of bleeding areas detection for the simulated bleeding scenario and the three craniotomy phase scenarios were 94.40%, 84.44%, 89.48%, and 90.46%. ConclusionThe contours of the neurosurgical craniotomy bleeding regions can be detected along with the bleeding areas. SignificanceIt is beneficial for neurosurgeons to identify the bleeding areas by sending prioritized alerts for bleeding events. Furthermore, it is valuable for a task-level medical robot designed for a neurosurgical procedure, such as craniotomy, or a high-level robot designed for an entire neurosurgery procedure.

Commentary: Beware of Optical Illusion of the Alar Base in Unilateral Cleft Lip Nasal Deformity.

Sir: We would like to congratulate the authors on an important study that clarifies dogma regarding the unilateral cleft lip nasal deformity.1 A common misperception is that the cleft-side alar base is displaced lateral, and that correction requires medial advancement (Fig. 1A). Just as we have, the authors provide objective evidence of the reverse2: relative to facial midline, the non-cleft alar base is displaced laterally and needs to be centralized (Fig. 1B). Mulliken has proposed that surrounding regional contours may lead to optical illusions of alar base position.3 We feel that the misperception may be due to human cognitive tendencies to reference the more natural appearing non-cleft side when explaining the less-familiar cleft-side contours. Regardless, a clear understanding of the deformity is critical to improving treatment and avoiding iatrogenic anomalies.Fig. 1.: A, Optical illusion. In the absence of facial regions outside the nasolabial region it may seem that correction of the deformity involves downward rotation of the medial lip philtrum and medial advancement of the lateral lip and alar base. B, Context. Facial features and reference lines make the actual displacements readily perceived. The non-cleft alar base and columella need to be centralized, and the cleft alar base needs to be advanced anteriorly. If the lateral lip and alar base were advanced medially, it would produce further imbalance, aggravate the nasal twist, and introduce iatrogenic deformity. Modified with permission from Plast Reconstr Surg 2020:145;185–199.Sakamoto et al analyzed a large sample of photographs and suggested that the cleft alar base is displaced medial relative to age-matched controls.1 Their findings differed from ours in that we found the cleft alar base to be retruded, but normal in medial-lateral position.2 Although they attributed the discrepancy to their larger sample size,1 there are several other explanations to consider. First, whereas our study examined standardized 3D images, they used 2D photographs that involve considerable error from lens distortions, parallax, and uncontrolled head rotation. The appearance of depth from a retruded cleft alar base on 2D images could be extrapolated as variation in medial-lateral position, and error from parallax may be magnified when images are tilted to normalize head position. Second, whereas we measured alar base position using calibrated images with accuracy to within 1 mm, they normalized dimensions on non-standardized photographs using inter-endocantion distance expressed as ratios. These ratios don’t account for the fact that the inter-endocanthion distance widens with increasing cleft severity.2 The findings that Sakamoto et al report may therefore simply be a changing denominator rather than true alteration of cleft alar base position. Although further analysis of well-controlled 3D images would be helpful, the medial position of cleft alar base relative to controls that they report is minimal and may not be of clinical relevance. Most figures depicting unilateral cleft lip nasal deformity crop out facial features beyond the nasolabial region and some research studies advocate that doing so reduces bias when rating esthetic outcomes.4 We are glad to see that Figure 4 in Sakamoto’s article1 is identical to Figure 1 in ours.2 The midline and the endocanthion verticals illustrate the medial-lateral displacements, whereas the alar base horizontals illustrate the relative vertical disparities. Visualization of the whole face allows balance to be judged, and the reference lines guide us to correct perceptions (Fig. 1). We propose that those references should be standard and caution against future research that crops out non-nasolabial regions without providing a mechanism to judge balance. Finally, based on Sakamoto’s,1 Fisher’s,5 and our2 studies, we propose that centralizing the columella and rebalancing the nasal base elements should be the fundamental principle in the correction of the unilateral cleft lip nasal deformity. DISCLOSURE The authors have no financial interest to declare in relation to the content of this article.

Artifact removal in the contour areas of SAXS-CT images by Tikhonov-L1 minimization.

Small-angle X-ray scattering (SAXS) coupled with computed tomography (CT), denoted SAXS-CT, has enabled the spatial distribution of the characteristic parameters (e.g. size, shape, surface, length) of nanoscale structures inside samples to be visualized. In this work, a new scheme with Tikhonov regularization was developed to remove the effects of artifacts caused by streak scattering originating from the reflection of the incident beam in the contour regions of the sample. The noise due to streak scattering was successfully removed from the sinogram image and hence the CT image could be reconstructed free from artifacts in the contour regions.

Research on machining parameter optimization in finishing milling with multiple constraints

High efficiency and high precision milling, as the eternal goal of CNC machining, needs to balance many constraints for selecting the most reasonable processing parameters. This paper presents an efficient machining parameter optimization method for finishing milling operation with multiple constraints. Firstly, under the multiple constraints of parameter feasible region, milling force, milling stability, roughness, and machining contour accuracy, a multi-variable parameter optimization model with machining efficiency as the objective is established. A four level cycle optimization strategy has been detailly described for solving the optimization problem, in which the feed per tooth is optimized by using the golden section method, and with the aid of the random vector search method, the spindle speed, radial, and axial depth cuts are both numerically iterated. The optimal machining parameter combination of the tooth number, feed per tooth, spindle speed, radial, and axial depth of cuts are achieved at last. Finally, the experimental verification results show that the proposed method can greatly improve the machining efficiency under chatter free condition and achieve an efficient finishing milling with consideration of the multiple constraints.

COVID-19 Volumetric Pulmonary Lesion Estimation on CT Images using a U-NET and Probabilistic Active Contour Segmentation.

A two-step method for obtaining a volumetric estimation of COVID-19 related lesion from CT images is proposed. The first step consists in applying a U-NET convolutional neural network to provide a segmentation of the lung-parenchyma. This architecture is trained and validated using the Thoracic Volume and Pleural Effusion Segmentations in Diseased Lungs for Benchmarking Chest CT Processing Pipelines (PleThora) dataset, which is publicly available. The second step consists in obtaining the volumetric lesion estimation using an automatic algorithm based on a probabilistic active contour (PACO) region delimitation approach. Our pipeline successfully segmented COVID-19 related lesions in CT images, with exception of some mislabeled regions including lung airways and vasculature. Our workflow was applied to images in a cohort of 50 patients.

Evaluating Clinical Acceptability of Organs-at-Risk Segmentation in Head & Neck Cancer (HNC) by Open-Source 3D Convolutional Neural Networks (CNNs)

<h3>Purpose/Objective(s)</h3> To evaluate the performance and clinical acceptability of the highest objective performing 3D convolutional neural network models developed to contour regions of interest (ROIs) in medical images. <h3>Materials/Methods</h3> Radiation treatment planning data from 582 HNC patients treated at a large tertiary cancer center were used for this study. Eleven open-source 3D segmentation models (6 UNETs, 3 DenseNets, 1 3D pyramid abstraction network, and 1 3D recurrent slice-wise attention network), originally engineered for medical image segmentation, were trained to segment 19 OARs. The same 80/10/10 split was used for training, validating and testing corresponding to 479, 44 and 59 scans. A model's performance was ranked by volumetric Dice similarity coefficient (DSC) and Hausdorff distance (HD). The best performing model was selected and re-trained on full-resolution patient scans and validated using sliding-window inference. With the best model, two radiation oncologists assessed clinical acceptability, using 5-scale ratings of 5 for "anatomically perfect, not requiring editing" and score 1 for "anatomically incorrect, unusable for planning purpose." At least 10 cases were randomly selected and deep learning-based (DLCs) and manual contours (MCs), for the entire range of the OAR, were presented to the blinded observers. The mean ratings between DLC and MC were compared using Wilcoxon signed rank test and Student's T-test for the groups where significant differences between the groups represented the ability to distinguish between the MC and DLC contours by the observers. <h3>Results</h3> A simple 3D Unet performed the best among the 11 models when comparing average DSC (0.75 ± 0.12) and HD (mean HD 0.44 ± 0.09) for every OAR. Mandible (0.90 ± 0.03), brainstem (0.89 ± 0.04), and spinal cord (0.85 ± 0.05) received the highest mean DSC, and chiasm (0.38 ± 0.19), optic nerves (0.65 ± 0.15), and lips (0.70 ± 0.09) the lowest mean DSC. Lens (0.052 ± 0.019), optic nerve (0.069 ± 0.026), and chiasm (0.075 ± 0.02) received the lowest mean HD, and spinal cord (1.15 ± 0.15), mandible (1.11 ± 0.18), and brachial plexus (0.90 ± 0.12) the highest mean HD. For clinical acceptability, MC received a significantly higher rating than DLC (3.75 ± 0.77 vs. 3.23 ± 0.86) when all OARs were considered (<i>P</i> < 0.01). When evaluating OARs individually, MC showed significantly higher ratings for brainstem, esophagus, larynx, eyes, optic nerves, while lips, whereas parotids, acoustics, and lenses were indistinguishable. <h3>Conclusion</h3> Simple 3D architectures consistently outcompete more complex networks by quantitative measures. Qualitative assessment for clinical acceptability may not agree with quantitative performance, especially when the entire range of OARs is evaluated.

A Magic Wand Selection Tool for Surface of 3D Model

Introduction: segmentation of 3d shapes is a fundamental problem in computer graphics and computer-aided design. It has received much plays attention in recent years. The analysis and research methods of 3d mesh models have established reliable mathematical foundations in graphics and geometric modeling. Compared with color and texture, shape features describe the shape information of objects from geometric structure features. And it an important role in a wide range of applications, including mesh parameterization, skeleton extraction, resolution modeling, shape retrieval, character recognitio,, robot navigation, and many others. Methods: The interactive selection surface of models is mainly used for shape segmentation. The common method is boundary-based selection, which requires user input some stokes near the edge of the selected or segmented region. Chen et al.introduced an approach to join the specified points form the boundaries for region segmentation on the surface. Funkhouser et al. improve the Dijkstra algorithm to find segmentation boundary contour. The graph cut algorithm use the distance between the surface and its convex hull as the growing criteria to decompose a shape into meaningful components. The watershed algorithm, widely used for image segmentation, is a region-growing algorithm with multiple seed points. Wu and Levine use simulated electrical charge distributions over the mesh to deal with the 3D part segmentation problem. Other methods using a watershed algorithm for surface decomposition. Results: We implemented our algorithm in C++ and Open MP and conducted the experiments on a PC with a 3.07 GHz Intel(R) Core(TM) i7 CPU and 6 GB memory. Our method can get a similar region under different interaction vertices in specific regions. Figure 6a and Figure 6b are the calculation results of tolerance region selection of this algorithm in a certain region of kitten model at two different interaction points, from which we can see the obtained regions are similar from different vertices in this region. Figure 6c and figure 6d are two different interactive points in the same region, and the region selection results are obtained by Region growing technique. Discussion: In this paper, we proposed a novel magic wand selection tool to interactive select surface of 3D model. The feature vector is constructed by extracting the HKS feature descriptor and means curvature of 3D model surface, which allows users to input the feature tolerance value for region selection and improves the self-interaction of users. Many experiments show that our algorithm has obvious advantages in speed and effectiveness. The interactive generation of region boundary is very useful for many applications including model segmentation. Conclusion: In consideration of a couple of requirements including user-friendliness and effectiveness in model region selection, we propose a novel magic wand selection tool to interactive selection surface of 3D models. First, we pre-compute the heat kernel feature and mean curvature of the surface, and then form the eigenvector of the model. Then we provide two ways for region selection. One is to select the region according to the feature of tolerance value. The other is to select the region that aligns with stroke automatically. Finally, we use the geometry optimization approach to improve the performance of the computing region con-tours. Extensive experimental results show that our algorithm is efficient and effective.

Free rectus abdominis myocutaneous flap for the reconstruction of major and complex defects in the craniofacial regions after ablation of advanced sinonasal carcinomas

Objective: To explore the method and plausibility of using free rectus abdominis flap (FRAF) to reconstruct the major and complex defects in the craniofacial regions after ablation of advanced sinonasal carcinoma. Methods: From 2007 to 2018 at No. 980 Hospital of Joint Logistic Supportive Force of People's Liberation Army of China, 13 patients with advanced carcinoma of nasal cavity and paranasal sinuses, including 11 males and 2 females, aged from 33 to 67 years, were treated with FRAF to repair the invasion of skull base, face and orbit. Based on adequate and meticulous preoperative evaluations on patients and tumors, complete resection of tumor bulks was performed. According to the sites, characters and extents of the defects, FRAF was introduced in different ways into reconstruction of major and complex defects in the craniofacial regions after resection of advanced sinonasal carcinomas, restoring the structure and contour of the craniofacial region. Results: Complete resection of tumors was achieved in all cases. For repairing the major and complex defects resulted from tumor ablation, FRAF graft was conducted in 13 patients with advanced sinonasal carcinomas. The recipient vessels were facial artery and vein in 9 cases, superficial temporal artery and vein in 4 cases. Vascular bridging with the external jugular vein was carried out in 2 cases who underwent submandibular neck dissections, in which facial artery and vein were used as recipient vessels. Free fibular flap was used along with FRAF in a case undergone bilateral maxillectomy. Pedicled galea capitis and periosteal flap were applied simultaneously in 3 cases with dura defects at the skullbase resulted from tumor resection. All flaps used in 13 cases survived uneventfully without major complications, with successful repair of defects, reconstruction of structures and restoration of contour of craniofacial regions. Conclusion: With adequate evaluation of resectability of tumors as well as nature and extent of defects after tumor removal, FRAF transplantation is an ideal method and strategy of choice for reconstructing the major and complex defects in craniofacial regions resulted from tumor ablation of advanced sinonasal carcinomas.