Search

Breast-region segmentation is an important step for density estimation and Computer-Aided Diagnosis (CAD) systems in Magnetic Resonance Imaging (MRI). Detection of breast-chest wall boundary is often a difficult task due to similarity between gray-level values of fibroglandular tissue and pectoral muscle. This paper proposes a robust breast-region segmentation method which is applicable for both complex cases with fibroglandular tissue connected to the pectoral muscle, and simple cases with high contrast boundaries. We present a decision-making framework based on geometric features and support vector machine (SVM) to classify breasts in two main groups, complex and simple. For complex cases, breast segmentation is done using a combination of intensity-based and atlas-based techniques; however, only intensity-based operation is employed for simple cases. A novel atlas-based method, that is called localized-atlas, accomplishes the processes of atlas construction and registration based on the region of interest (ROI). Atlas-based segmentation is performed by relying on the chest wall template. Our approach is validated using a dataset of 210 cases. Based on similarity between automatic and manual segmentation results, the proposed method achieves Dice similarity coefficient, Jaccard coefficient, total overlap, false negative, and false positive values of 96.3, 92.9, 97.4, 2.61 and 4.77%, respectively. The localization error of the breast-chest wall boundary is 1.97 mm, in terms of averaged deviation distance. The achieved results prove that the suggested framework performs the breast segmentation with negligible errors and efficient computational time for different breasts from the viewpoints of size, shape, and density pattern. [ABSTRACT FROM AUTHOR]

In the marketing area, new trends are emerging, as customers are not only interested in the quality of the products or delivered services, but also in a stimulating shopping experience. Creating and influencing customers' experiences has become a valuable differentiation strategy for retailers. Therefore, understanding and assessing the customers' emotional response in relation to products/services represents an important asset. The purpose of this paper consists of investigating whether the customer's facial expressions shown during product appreciation are positive or negative and also which types of emotions are related to product appreciation. We collected a database of emotional facial expressions, by presenting a set of forty product related pictures to a number of test subjects. Next, we analysed the obtained facial expressions, by extracting both geometric and appearance features. Furthermore, we modeled them both in an unsupervised and supervised manner. Clustering techniques proved to be efficient at differentiating between positive and negative facial expressions in 78% of the cases. Next, we performed more refined analysis of the different types of emotions, by employing different classification methods and we achieved 84% accuracy for seven emotional classes and 95% for the positive vs. negative. [ABSTRACT FROM AUTHOR]

Texture feature plays a predominant role in recognizing face images. However different persons can have similar texture features that may degrade the system performance. Hence in this paper, the problem of face similarity is addressed by proposing a solution which combines textural and geometrical features. An algorithm is proposed to combine these two features. Five texture descriptors and few geometrical features are considered to validate the proposed system. Performance evaluations of these features are carried out independently and jointly for three different issues such as expression variation, illumination variation and partial occlusion with objects. It is observed that the combination of textural and geometrical features enhance the accuracy of face recognition. Experimental results on Japanese Female Facial Expression (JAFFE) and ESSEX databases indicate that the texture descriptor Local Binary Pattern achieves better recognition accuracy for all the issues considered.

Identification of functionally connected regions while at rest has been at the forefront of research focusing on understanding interactions between different brain regions. Studies have utilized a variety of approaches including seed based as well as data-driven approaches to identifying such networks. Most such techniques involve differentiating groups based on group mean measures. There has been little work focused on differences in spatial characteristics of resting fMRI data. We present a method to identify between group differences in the variability in the cluster characteristics of network regions within components estimated via independent vector analysis (IVA). IVA is a blind source separation approach shown to perform well in capturing individual subject variability within a group model. We evaluate performance of the approach using simulations and then apply to a relatively large schizophrenia data set (82 schizophrenia patients and 89 healthy controls). We postulate that group differences in the intra-network distributional characteristics of resting state network voxel intensities might indirectly capture important distinctions between the brain function of healthy and clinical populations. Results demonstrate that specific areas of the brain, superior and middle temporal gyrus that are involved in language and recognition of emotions, show greater component level variance in amplitude weights for schizophrenia patients than healthy controls. Statistically significant correlation between component level spatial variance and component volume was observed in 19 of the 27 non-artifactual components implying an evident relationship between the two parameters. Additionally, the greater spread in the distance of the cluster peak of a component from the centroid in schizophrenia patients compared to healthy controls was observed for seven components. These results indicate that there is hidden potential in exploring variance and possibly higher-order measures in resting state networks to better understand diseases such as schizophrenia. It furthers comprehension of how spatial characteristics can highlight previously unexplored differences between populations such as schizophrenia patients and healthy controls.

Due to the close resemblance between overlapping and cancerous nuclei, the misinterpretation of overlapping nuclei can affect the final decision of cancer cell detection. Thus, it is essential to detect overlapping nuclei and distinguish them from single ones for subsequent quantitative analyses. This paper presents a method for the automated detection and classification of overlapping nuclei from single nuclei appearing in cytology pleural effusion (CPE) images. The proposed system is comprised of three steps: nuclei candidate extraction, dominant feature extraction, and classification of single and overlapping nuclei. A maximum entropy thresholding method complemented by image enhancement and post-processing was employed for nuclei candidate extraction. For feature extraction, a new combination of 16 geometrical and 10 textural features was extracted from each nucleus region. A double-strategy random forest was performed as an ensemble feature selector to select the most relevant features, and an ensemble classifier to differentiate between overlapping nuclei and single ones using selected features. The proposed method was evaluated on 4000 nuclei from CPE images using various performance metrics. The results were 96.6% sensitivity, 98.7% specificity, 92.7% precision, 94.6% F1 score, 98.4% accuracy, 97.6% G-mean, and 99% area under curve. The computation time required to run the entire algorithm was just 5.17 s. The experiment results demonstrate that the proposed algorithm yields a superior performance to previous studies and other classifiers. The proposed algorithm can serve as a new supportive tool in the automated diagnosis of cancer cells from cytology images.

This paper presents a parametric study on fire-tube steam boilers’ heat exchanging unit using Entropy Generation Minimization (EGM) technique. Mathematical equation that models entropy generation rate (Ṡgen), and entropy generation number (Ns), were formulated. The entropy generation number was related to the geometric features of the system. Solving the equations showed the effects of each geometric feature on both pressure drop and Ns. Correlation equation relating dimensionless fire-tube length to dimensionless tube diameter was also obtained. The obtained correlation equation is a tool to be incorporated into fire-tube steam boiler design, so as to account for entropy generation during design.

The article deals with application of geometric features and algorithms for optimization of manufacturing processes. Algorithms allow to create strategies based on a properties of geometric features. These strategies can be used to remove repetitive tasks and shorting of time for creation a machining process. Article describes the use of a Feature Based Machining (FBM) and a recognition of geometric features. [ABSTRACT FROM AUTHOR]

To solve the problem that the virtual 3D cable of CREO can easily cause interference detection difficult because the cable has a flexible optional deformation characteristic, an algorithm that expresses the cable discretely by using dimensional coordinates point is put forward. Firstly, the algorithm expresses the cable entities by using cable centerline discrete spatial coordinates point. Secondly, space position relation of structure is analyzed. Finally, by utilizing the relationship of space position between cables as well as cable and structure, the fast interference detection for 3D cables is realized. The algorithm is validated on the platform Visual Studio 2010 and CREO by using C++ language and TOOLKIT tools. The results show that the algorithm is fast and has better accuracy to meet the actual needs of the project. [ABSTRACT FROM AUTHOR]

This paper presents the first method that enables the fully automatic generation of triangular meshes suitable for the so-called non-uniform rational B-spline (NURBS)-enhanced finite element method (NEFEM). The meshes generated with the proposed approach account for the computer-aided design boundary representation of the domain given by NURBS curves. The characteristic element size is completely independent of the geometric complexity and of the presence of very small geometric features. The proposed strategy allows to circumvent the time-consuming process of de-featuring complex geometric models before a finite element mesh suitable for the analysis can be produced. A generalisation of the original definition of a NEFEM element is also proposed, enabling to treat more complicated elements with an edge defined by several NURBS curves or more than one edge defined by different NURBS. Three examples of increasing difficulty demonstrate the applicability of the proposed approach and illustrate the advantages compared with those of traditional finite element mesh generators. Finally, a simulation of an electromagnetic scattering problem is considered to show the applicability of the generated meshes for finite element analysis. ©2016 The Authors. International Journal for Numerical Methods in Engineering published by John Wiley & Sons Ltd. [ABSTRACT FROM AUTHOR]

In ground-penetrating radar (GPR) subsurface target reconstruction, various techniques based on amplitude (or energy) information of echoes from metal target with small size can work well. However, for environmental and geological applications, the quantitative analysis of the target's geometric features, like location, shape and size, is exactly what we are concerned about. Amplitude-based reconstruction method faces challenges in these applications. A large sloped target under the surface may lead to abundant virtual image energy and cause position deviation. In addition, the echoes from the inner part of the penetrable dielectric target may be too weak to be detected. However, changes in phase may highlight the effects of echoes from the target on that from the surroundings, even if the effect is small due to the weak energy. In this paper, a novel method based on changes in phase is proposed to reconstruct subsurface large sloped dielectric target. To remove the virtual image, the main idea is based on the geometric relationship between the recorded signal plotted beneath the receiving antenna and the reflected signal emanated from the target position which is “ahead” or “behind” of the receiving antenna. Furthermore, the electromagnetic (EM) wave propagating through the penetrable target will change its velocity and result in advancing or lagging related to the geometric shape of the target. In this case, the weak echoes from the back surface of the target can be compensated according to the advancing or lagging. With the proposed method, the virtual image is eliminated and both front and back surface of the target are reconstructed. Results from the laboratory experiments demonstrate the validity of the proposed method. [ABSTRACT FROM AUTHOR]

Copyright of Journal of Signal Processing is the property of Journal of Signal Processing and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Copyright of Scientific Bulletin of National Mining University is the property of National Mining University, State Higher Educational Institution and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use. This abstract may be abridged. No warranty is given about the accuracy of the copy. Users should refer to the original published version of the material for the full abstract. (Copyright applies to all Abstracts.)

Unsignalized intersections namely two-way stop-controlled intersection (TWSC) and all-way stop-controlled intersection (AWSC) are widely used in Akure. Five intersections consisting of three Tee and two Cross that were critical to traffic flow in the study area were selected for study. Data on geometric features were collected using odometer, while traffic parameters were captured and metered using cine camera placed at vantage positions from the intersections during peak and off-peak periods on week days. Traffic flows at the intersections were expressed as functions of traffic characteristics and geometric features of the approaches; while the effect of distances of intersections before and after the intersections studied were also incorporated as correction factors in the models. The models were developed using multiple linear regression technique with the aid of SPSS software and validated with empirical data other than those used for model calibration. Adjusted R2 values of 0.881 and 0.882 were obtained for Tee and Cross intersections respectively for peak period, while 0.938 and 0.940 respectively were obtained for off-peak period. These indicate that the flow models are very robust in replicating the observed data. The predictive models have the potential to accurately estimate traffic flow at intersections in the study area and other cities of the world with similar traffic conditions. [ABSTRACT FROM AUTHOR]

This paper presents a multimodal emotion recognition method that uses a feature-level combination of three-dimensional (3D) geometric features (coordinates, distance and angle of joints), kinematic features such as velocity and displacement of joints, and features extracted from daily behavioral patterns such as frequency of head nod, hand wave, and body gestures that represent specific emotions. Head, face, hand, body, and speech data were captured from 15 participants using an infrared sensor (Microsoft Kinect). The 3D geometric and kinematic features were developed using raw feature data from the visual channel. Human emotional behavior-based features were developed using inter-annotator agreement and commonly observed expressions, movements and postures associated to specific emotions. The features from each modality and the behavioral pattern-based features (head shake, arm retraction, body forward movement depicting anger) were combined to train the multimodal classifier for the emotion recognition system. The classifier was trained using 10-fold cross validation and support vector machine (SVM) to predict six basic emotions. The results showed improvement in emotion recognition accuracy (The precision increased by 3.28% and the recall rate by 3.17%) when the 3D geometric, kinematic, and human behavioral pattern-based features were combined for multimodal emotion recognition using supervised classification.

Head pose estimation (HPE) is important in human–machine interfaces. However, various illumination, occlusion, low image resolution and wide scene make the estimation task difficult. Hence, a Dirichlet-tree distribution enhanced Random Forests approach (D-RF) is proposed in this paper to estimate head pose efficiently and robustly in unconstrained environment. First, positive/negative facial patch is classified to eliminate influence of noise and occlusion. Then, the D-RF is proposed to estimate the head pose in a coarse-to-fine way using more powerful combined texture and geometric features of the classified positive patches. Furthermore, multiple probabilistic models have been learned in the leaves of the D-RF and a composite weighted voting method is introduced to improve the discrimination capability of the approach. Experiments have been done on three standard databases including two public databases and our lab database with head pose spanning from −90° to 90° in vertical and horizontal directions under various conditions, the average accuracy rate reaches 76.2% with 25 classes. The proposed approach has also been evaluated with the low resolution database collected from an overhead camera in a classroom, the average accuracy rate reaches 80.5% with 15 classes. The encouraging results suggest a strong potential for head pose and attention estimation in unconstrained environment. [ABSTRACT FROM AUTHOR]

The object present in an image is an important content and can be used in CBIR applications. Identifying and representing the shape of the object is indeed complex due to the uncertainties in the boundary of the object of interest. In this paper, we have proposed Fuzzy-Object-Shape to capture the shape of the object of interest along with the degree of impreciseness in the boundary information. The Fuzzy-Object-Shape information is extracted from each object in an image. This information provides a measure of closeness of the object of interest with well-known shapes. For each object, the fuzzy membership values are calculated and considered as feature vector. A similarity measure is proposed for measuring the degree of closeness of objects present in both query and database images. The performance of the proposed approach is compared with some of the recently proposed similar approaches. Benchmark dataset and uncontrolled dataset are used for the experiments and the performance of the proposed approach is encouraging. [ABSTRACT FROM AUTHOR]

We are studying image-based fish identification. Most traditional approaches use a fish image wherein extraction is easy given that the fish region contrasts with a white or uniform background. This research introduces an approach to give several feature points by manual operation. In our proposed approach, we are able to accept fish image with complicated background, including rocky area. Further, to investigate efficient features for fish recognition in image, we define various features, including geometric features, bag of visual words (BoVW) models, and texture features. We collected images comprising 129 fish species under various photographic conditions and applied our proposed method to these images. From our results, we confirmed that a combination of geometric features and BoVW models obtained the highest recognition accuracy. [ABSTRACT FROM AUTHOR]

This paper presents a parametric study on fire-tube steam boilers' heat exchanging unit using Entropy Generation Minimization (EGM) technique. Mathematical equation that models entropy generation rate (Sgen), and entropy generation number (Ns), were formulated. The entropy generation number was related to the geometric features of the system. Solving the equations showed the effects of each geometric feature on both pressure drop and Ns. Correlation equation relating dimensionless fire-tube length to dimensionless tube diameter was also obtained. The obtained correlation equation is a tool to be incorporated into fire-tube steam boiler design, so as to account for entropy generation during design. [ABSTRACT FROM AUTHOR]

Nowadays, selective laser melting (SLM) technology features is somehow unknown mainly on the process parameters needed for each material. The main goal of this work is to give better process parameters combination when CoCrMo powders are utilized. A statistical methodology based on the experimentation has been applied. In this study, a full factorial design is used. It assesses the impact of process parameters on surface quality (Q), flatness, overlapping, and surface roughness. The distinguishing feature is the melting of single layers in variable layers powder with a continuous laser. Based on these results combined with visual observation of the solidified tracks, the effect of process parameters on the surface roughness for each case is investigated. The results show that surface roughness (Ra) is affected strongly by laser power, scan spacing, and scanning speed thereby giving an average roughness at 8 μm. In addition, other roughness parameters such as the peak height (Rp) and depth of the lowest point of the profile (Rv) are found to be useful tools for evaluation of top quality. The best layers have been showed when the laser powers are 300 and 400 W and the scanning space of 450 and 600 μm. These layers are classified with (Q) less than 2. [ABSTRACT FROM AUTHOR]

We present a method for the automatic localization of facial landmarks that integrates nonrigid deformation with the ability to handle missing points. The algorithm generates sets of candidate locations from feature detectors and performs combinatorial search constrained by a flexible shape model. A key assumption of our approach is that for some landmarks there might not be an accurate candidate in the input set. This is tackled by detecting partial subsets of landmarks and inferring those that are missing, so that the probability of the flexible model is maximized. The ability of the model to work with incomplete information makes it possible to limit the number of candidates that need to be retained, drastically reducing the number of combinations to be tested with respect to the alternative of trying to always detect the complete set of landmarks. We demonstrate the accuracy of the proposed method in the face recognition grand challenge database, where we obtain average errors of approximately 3.5 mm when targeting 14 prominent facial landmarks. For the majority of these our method produces the most accurate results reported to date in this database. Handling of occlusions and surfaces with missing parts is demonstrated with tests on the Bosphorus database, where we achieve an overall error of 4.81 and 4.25 mm for data with and without occlusions, respectively. To investigate potential limits in the accuracy that could be reached, we also report experiments on a database of 144 facial scans acquired in the context of clinical research, with manual annotations performed by experts, where we obtain an overall error of 2.3 mm, with averages per landmark below 3.4 mm for all 14 targeted points and within 2 mm for half of them. The coordinates of automatically located landmarks are made available on-line. [ABSTRACT FROM PUBLISHER]

In this paper, we propose a novel method including segmentation, a combination of new and well-known feature extraction and classification methods to classify plant leaves. The aim of the proposed features is to distinguish leaf margins, which cannot be distinguished using commonly used geometric features. Additionally, Linear Discriminant Classifier is used for classification, therefore using features that are noisy for some leaf types does not reduce the performance of the system. The proposed system outperforms the well-known geometric methods that are used for leaf classification. [ABSTRACT FROM AUTHOR]

The Internet of Medical Things (IoMT) combines the Internet of Things (IoT) with medical equipment to provide better patient comfort, cost-effective medical solutions, faster hospital treatments, and even more individualized healthcare. Brain tumors are caused by a mass of random cells in the brain, harming the brain and posing a risk. Brain image recognition is a difficult task these days. This paper explains the application of a support value based deep neural network (SDNN) in e-Health care as use of the Internet of Medical Things (IoMT) innovation for the early detection and accurate classification of cancerous cells in brain pictures. Initially, as an exploration database, photos based on IoT innovation and clinical images are collected. The input brain picture is stripped of its skull for brain area extraction during the preprocessing stage. The effective features such as entropy, geometric, and texture features are retrieved from the preprocessed output images. Finally, the suggested support value based adaptive deep neural network (SDNN) recognition distinguishes between normal and abnormal brain images into normal or abnormal images dependent on the extracted features. Experimental results show that our proposed SDNN approach accomplishes an accuracy as high as 94.30%. In the meantime, the other existing methods, CNN+ReLU is 91.02%, CNN+PReLU, CNN+BN+ReLU, CNN+BN+PReLU achieve had the worst accuracy of 82.05, 85.55, and 83.6% independently. Compared with other existing works, our proposed methodology achieves higher outcomes. [ABSTRACT FROM AUTHOR]

A simple method is proposed for the derivation of bit error floor (BEF) expressions of multi‐level phase modulations with noisy phase reference based on the geometric features of constellations. To validate the result, average BEF expressions of 8 + 8 and 4 + 12 amplitude and phase‐shift keyings adopted in digital video broadcasting‐satellite 2 extensions are derived and consultative committee for space data systems, and analyse the results. [ABSTRACT FROM AUTHOR]

Node localization in two-dimensional (2D) and three-dimensional (3D) space for wireless sensor networks (WSNs) remains a hot research topic. To improve the localization accuracy and applicability, we first propose a quantum annealing bat algorithm (QABA) for node localization in WSNs. QABA incorporates quantum evolution and annealing strategy into the framework of the bat algorithm to improve local and global search capabilities, achieve search balance with the aid of tournament and natural selection, and finally converge to the best optimized value. Additionally, we use trilateral localization and geometric feature principles to design 2D (QABA-2D) and 3D (QABA-3D) node localization algorithms optimized with QABA, respectively. Simulation results show that, compared with other heuristic algorithms, the convergence speed and solution accuracy of QABA are greatly improved, with the highest average error of QABA-2D reduced by 90.35% and the lowest by 17.22%, and the highest average error of QABA-3D reduced by 75.26% and the lowest by 7.79%.

The design of machine parts of different sizes and shapes has become relevant in the manufacturing industry which requires five-axis machine tools of high dynamic performance; different machining test pieces have been used to test and reflect the machine tool's performance. The S-shaped is still under development and consideration of which a new test piece better than the S-shaped part has been recommended to be put forward making the NAS979 the only standardized test piece; however, it has some limitations. Hence, this study proposes a new test piece to objectively satisfy the demand for machine tools with higher dynamic performance, which shows much improvement over the standard NAS979 and is the best alternative to the S-shaped test piece, and it combines the geometric and kinematic features of both test pieces. Geometrically, it has non-uniform surface continuity, variable twist angle, and variable curvature; and the cutting tool moves in close and opened angles along the tool path; there is sudden rise and fall of axes' velocity, acceleration, and jerk with much impact during machining which makes the S-cone test piece be machined by only five-axis machine tools with high dynamic performance, and has a better dynamic performance identification effect than the S-shaped test piece based on the trajectory test. Detailed work on the validation of the machine tool's dynamic performance using the S-cone part will be captured next part of this study., Competing Interests: Competing interestsThe authors declare no competing interests., (© The Author(s), under exclusive licence to Springer-Verlag London Ltd., part of Springer Nature 2023, Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.)

The paper offers a mathematical model for imaging blood vessels. The model relies on division of blood-vessel geometric characteristics into local parameters, which are calculated directly in the process of vessel tracing, and integral (global) features, which are used later for forming classification features. Methods and algorithms to evaluate geometric features of three-dimensional three structures are considered. The immunity of evaluation methods to various kinds of noise and possibility to cluster a blood-vessel sample are investigated experimentally. [ABSTRACT FROM AUTHOR]

A novel method is presented for hand shape identification based on abductive machine learning. We developed several models and investigated their performance on raw hand shape data for 20 and 40 participants in the form of three different categories of geometric measurements: twelve finger features, two palm features, and three whole hand features. Performance was compared when using each category of features separately and when combining them together. Moreover, we describe two novel and more effective approaches using an ensemble of three abductive networks combined at either the score level or the decision level. The effect of doubling the number of participants from 20 to 40 was studied as well. The ensemble approach achieved overall identification accuracies of 100 and 98.3333 % for the 20-participant and 40-participant datasets, respectively. This compares favorably with other learning approaches tried on the same datasets, including decision trees, support vector machines, and rule-based classifiers. [ABSTRACT FROM AUTHOR]

Face is one of the most important biometric of human and contains lots of useful information such as gender, age, race and identity. Gender classification is very easy for human but it considers a challenge for computers. Gender classification through face images has recently been considered so much. Gender recognition can be useful in interaction between human and computer like identifying individual's identity. It is also applicable in TV networks in order to study the rate of viewers. Various algorithms have been designed for this issue and each of them has unraveled that to some extent. The last obtained rate to identify gender was through article written by Mozaffari who obtained mean rate of 83% for identification. It is the proposed method of the present study which has brought identification rate to 92.5. in this method we draw out face features based on Gabor filters and local binary patterns. These features are resistant against noise and they select proper features against bottleneck of images. In order to obtain a proper classification, we use self-organized map (SOM) (type of artificial neural network). This neural network finds the proper weights for each gender with very little error. Obtained results are compared with existing datasets and therefore, superiority of the proposed method would be evident. [ABSTRACT FROM AUTHOR]

The segmentation of touching characters is still a challenging task, posing a bottleneck for offline Chinese handwriting recognition. In this paper, we propose an effective over-segmentation method with learning-based filtering using geometric features for single-touching Chinese handwriting. First, we detect candidate cuts by skeleton and contour analysis to guarantee a high recall rate of character separation. A filter is designed by supervised learning and used to prune implausible cuts to improve the precision. Since the segmentation rules and features are independent of the string length, the proposed method can deal with touching strings with more than two characters. The proposed method is evaluated on both the character segmentation task and the text line recognition task. The results on two large databases demonstrate the superiority of the proposed method in dealing with single-touching Chinese handwriting. [ABSTRACT FROM AUTHOR]

As the technology of high-precision 3D laser scanning becomes increasingly prevalent in the fields of hydraulic building modeling and deformation monitoring, the quality of point clouds plays an increasingly crucial role in data processing. This paper investigates an automatic extraction method of anchor rod points based on geometric features, which focuses on the influence of anchor rod points and mixed pixels in the data of an underground powerhouse of a pumped storage power station on modeling and deformation monitoring during the construction period. This workflow consists of two steps that can automatically extract anchor rod points from high-density point cloud data. Triangular mesh features in the local neighborhood and the parameters of the anchor rods are used to locate the anchor rod in downsampled data, and curvature features are used to extract anchor rod points precisely. The experiment of extracting anchor rods shows that the accuracy of this method of initial identification is 97.2%. Furthermore, precise extraction based on curvature curve fitting is applicable. This method can accurately separate the three types of anchor rods from the dense point cloud on the rough surface of a cavern roof; the false-extraction rate of anchor rod points is about 0.11% to 5.09%. This method can provide high-quality and dependable data sources for the precise registration, modeling and deformation analysis of point clouds in a construction cavern.

Hemodialysis is an important means to maintain life in patients with end-stage renal disease (ESRD). Approximately 76.8% of patients who begin hemodialysis do so through catheters, which play vital roles in the delivery of hemodialysis to patients. During the past decade, the materials, structures, and surface-coating technologies of catheters have constantly been evolving to ameliorate catheter-related problems, such as recirculation, thrombosis, catheter-related infections, and malfunction. In this study, based on the commercial GDK catheter, six catheter models (GDK, GDK1, GDK2, GDK3, GDK4, and GDK5) with different lumen diameters and different geometric features of side holes were established, and computational flow dynamics (CFD) were used to measure flow rate, shear stress, residence time (RT), and platelet lysis index (PLI). These six catheters were then printed with polycarbonate PC using 3D printing technology to verify recirculation rates. The results indicated that: (1) the catheter with a 5.5 mm outer diameter had the smallest average shear stress in the arterial lumen and the smallest proportion of areas with shear stress > 10 pa. With increasing catheter diameter, the shear stress in the tip volume became lower, the average RT increased, and the PLI decreased due to larger changes in shear stress; (2) the catheters with oval-shaped side holes had smaller shear stress levels than those with circular-shaped holes, indicating that the oval design was more effective; (3) the catheter with parallel dual side holes had uniformly distributed flow around side holes and exhibited lower recirculation rates in both forward and reverse connections, while linear multi-side holes had higher shear stress levels due to the large differences in flow around side holes. The selection of the material and the optimization of the side holes of catheters have significant impacts on hemodynamic performances and reduce the probability of thrombosis, thus improving the efficiency of dialysis. This study would provide some guidance for optimizing catheter structures and help toward the commercialization of more efficient HD catheters.

The type and ratio of abnormal red blood cells (RBCs) in blood can be identified through peripheral blood smear test. Accurate classification is important because the accompanying diseases indicated by abnormal RBCs vary. In clinical practice, this task is time-consuming because the RBCs are manually classified. In addition, because the classification depends on the subjective criteria of pathologists, objective classification is difficult to achieve. In this paper, an automatic classification method that is solely based on images of RBCs captured under a microscope and processed using machine learning (ML) is proposed. The size and hemoglobin abnormalities of RBCs were classified by optimizing the criteria used in clinical practice. For morphologically abnormal RBCs classification, used seven geometric features information (major axis, minor axis, ratio of major and minor axis, perimeter, circularity, number of convex hulls, difference between area and convex area) and five types of multiple classifiers (Support Vector Machine, Decision Tree, K-Nearest Neighbor, Random Forest, and Adaboost models). Among was categorized using SVM, highly accurate results (99.9%) were obtained. The classification is performed simultaneously, and results are provided to the user through a graphical user interface (GUI)., (© 2022 Wiley Periodicals LLC.)