Search

Background: Existing studies have shown that the brain network of major depression disorder (MDD) has abnormal topologies. However, constructing reliable MDD brain networks is still an open problem. New method: This paper proposed a reliable MDD brain network construction method. First, seven connectivity methods are used to calculate the correlation between channels and obtain the functional connectivity matrix. Then, the matrix is binarized using four binarization methods to obtain the EEG brain network. Besides, we proposed an improved binarization method based on the criterion of maximizing differences between groups: the adaptive threshold (AT) method. The AT can automatically set the optimal binarization threshold and overcome the artificial influence of traditional methods. After that, several network metrics are extracted from the brain network to analyze inter-group differences. Finally, we used statistical analysis and Fscore values to compare the performance of different methods and establish the most reliable method for brain network construction. Results: In theta, alpha, and total frequency bands, the clustering coefficient, global efficiency, local efficiency, and degree of the MDD brain network decrease, and the path length of the MDD brain network increases. Comparison with existing methods: The results show that AT outperforms the existing binarization methods. Compared with other methods, the brain network construction method based on phase-locked value (PLV) and AT has better reliability. Conclusions: MDD has brain dysfunction, particularly in the frontal and temporal lobes.

A fast recovery from disruptions is of vital importance for the reliability and sustainability of urban zone transit systems. It’s a complex task to coordinate multiple transit departments to mitigate disruption. There are many ways to response but it’s not always obvious how to combine them in an optimized manner. This study presents a new attempt to tackle this problem in a comprehensive and hierarchical way. At phase (i), a network-scale strategy selection optimization model is formulated as a joint routing and resource allocation (nJRRA) problem. This model produces solutions for efficient allocation of network resources to facilitate inter-department coordination. By constraining the problem further into an $\epsilon $ -constrained nJRRA problem, classic solution algorithms can be applied to solve the quadratically constrained quadratic program (QCQP). On top of this “basic model”, we propose adding a decision to delay the resource allocation decisions up to a maximum initiation time when the incident duration is stochastic. To test the models, a quasi-dynamic evaluation program with a given incident duration distribution is constructed using discretized time steps and discrete distributions. Five different demand patterns and four different disruption duration distributions (20 combinations) are tested on a small transit network. The results show that the two models outperform benchmark strategies such as using only line level adjustment or only bus bridging. They also highlight conditions when delaying the decision is preferred.

Background: The single-channel sleep EEG has the advantages of convenient collection, high-cost performance, and easy daily use, and it has been widely used in the classification of sleep stages. Methods: This paper proposes a single-channel sleep EEG classification method based on long short-term memory and a hidden Markov model (LSTM-HMM). First, the single-channel EEG is decomposed using wavelet transform (WT), and multi-domain features are extracted from the component signals to characterize the EEG characteristics fully. Considering the temporal nature of sleep stage changes, this paper uses a multi-step time series as the input for the model. After that, the multi-step time series features are input into the LSTM. Finally, the HMM improves the classification results, and the final prediction results are obtained. Results: A complete experiment was conducted on the Sleep-EDFx dataset. The results show that the proposed method can extract deep information from EEG and make full use of the sleep stage transition rule. The proposed method shows the best performance in single-channel sleep EEG classification; the accuracy, macro average F1 score, and kappa are 82.71%, 0.75, and 0.76, respectively. Conclusions: The proposed method can realize single-channel sleep EEG classification and provide a reference for other EEG classifications.

While many neural networks have been proposed for hyperspectral image classification, current backbones cannot achieve accurate results due to the insufficient representation by scalar features and always cause a cumbersome calculation burden. To solve the problem, we propose the capsule attention network (CAN), which combines an activity vector with an attention mechanism to improve HSI classification. In particular, we consider two attention mechanisms to improve the effectiveness of the activity vectors. First, an attention-based feature extraction (AFE) module is proposed to preprocess the spectral-spatial features of HSI data, which effectively mines useful information before the generation of the activity vectors. Second, we propose a self-weighted mechanism (SWM) to distinguish the importance of different capsule convolutions, which enhances the representation of the primary activity vectors. Experiments on four well-known HSI datasets have shown our CAN surpasses state-of-the-art (SOTA) methods on three widely used metrics with a much lower computational burden.

BackgroundPersistent HR-HPV causes cervical cancer, exhibiting geographic variance. Europe/Americas have higher HPV16/18 rates, while Asia/Africa predominantly have non-16/18 HR-HPV. This study in Fujian, Asia, explores non-16/18 HR-HPV infections, assessing their epidemiology and cervical lesion association for targeted prevention.MethodsA total of 101,621 women undergoing HPV screening at a hospital in Fujian Province from 2013 to 2019 were included. HPV genotyping was performed. A subset of 11,666 HPV-positive women with available histopathology results were analyzed to characterize HPV genotype distribution across cervical diagnoses.ResultsIn 101,621 samples, 24.5% tested positive for HPV. Among these samples, 17.3% exhibited single infections, while 7.2% showed evidence of multiple infections. The predominant non-16/18 high-risk HPV types identified were HPV 52, 58, 53, 51, and 81. Single HPV infections accounted for 64.1% of all HPV-positive cases, with 71.4% of these being non-16/18 high-risk HPV infections. Age-related variations were observed in 11,666 HPV-positive patients with pathological results. Cancer patients were older. In the cancer group, HPV52 (21.8%) and HPV58 (18.6%) were the predominant types, followed by HPV33, HPV31, and HPV53. Compared to single HPV16/18 infection, non-16/18 HPV predominated in LSIL. Adjusted odds ratios (OR) for LSIL were elevated: multiple HPV16/18 (OR 2.18), multiple non-16/18 HR-HPV (OR 2.53), and multiple LR-HPV (OR 2.38). Notably, solitary HPV16/18 conferred higher odds for HSIL and cancer.ConclusionOur large-scale analysis in Fujian Province highlights HPV 52, 58, 53, 51, and 81 as predominant non-16/18 HR-HPV types. Multiple HPV poses increased LSIL risks, while solitary HPV16/18 elevates HSIL and cancer odds. These findings stress tailored cervical cancer prevention, highlighting specific HPV impacts on lesion severity and guiding region-specific strategies for optimal screening in Asia, emphasizing ongoing surveillance in the vaccination era.

Abstract To reduce the dimensionality of EEG features and improve classification accuracy, we propose an improved hybrid feature selection method for EEG feature selection. First, MIC is used to remove irrelevant features and redundant features to reduce the search space of the second stage. QPSO is then used to optimize the feature in the second stage to obtain the optimal feature subset. Considering that both dimensionality and classification accuracy affect the performance of feature subsets, we design a new fitness function. Moreover, we optimize the parameters of the classifier while optimizing the feature subset to improve the classification accuracy and reduce the running time of the algorithm. Finally, experiments were performed on EEG and UCI datasets and compared with five existing feature selection methods. The results show that the feature subsets obtained by the proposed method have low dimensionality, high classification accuracy, and low computational complexity, which validates the effectiveness of the proposed method.

Despite awareness of the importance and value of sustainability for organizations, how sustainable development can be achieved remains poorly understood. This article aims to address this issue by examining the role of vision communication in promoting organizational sustainability. Drawing on the Upper Echelon Theory (UET) and existing literature on sustainability, we propose that sustainable vision communication can enhance sustainability performance by promoting and facilitating organizational change. Furthermore, we argue that CEO’s dialectical leadership behavior can supplement the efforts of leaders in achieving sustainable development, thereby strengthening the aforementioned relationship. Through a field study of 130 CEOs and their 562 top management team members, we tested our model and found evidence to support the hypotheses. This study thus advances our understanding of the processes underlying organizational sustainability and highlights the crucial role of leadership in this regard. We also discuss the theoretical implications of our findings and outline potential avenues for future research.

ObjectiveInvestigate HPV types in cervical specimens, their correlation with p16 expression in lesions, and diagnostic value for cervical lesions. Enhance clinical diagnosis reliability.MethodsRetrospective cross-sectional study at Fujian Maternity and Child Health Hospital’s Cervical Disease Center (Jun 2019-Dec 2021). Patients with abnormal cervical screening underwent colposcopy and conization. Pathological diagnosis based on colposcopy, cervical biopsy, ECC, and conization. Analyzed HPV genotyping (18 HR-HPV, 5 LR-HPV) and p16 expression correlation. Statistical analysis used R software.Resultshe expression of p16 is significantly associated with the infection of high-risk HPV types, such as 16, 33, 52, and 58, with an increased risk of 1.4 times or higher (OR=1.91, 3.14, 1.40, and 1.78, respectively). The risk of p16 expression increased 4-fold for multiple high-risk HPV types [adjusted OR (95% CI) = 4 (2.92~5.5), P-value

As the scale and complexity of software for industrial equipment increases, the risk of defects being introduced into the software also increases. Over-the-air (OTA) technology eliminates software defects by regularly updating and maintaining the software equipment. However, downloading an upgraded file is time-consuming. The differential compression algorithm can rapidly complete OTA software upgrades. To solve the low-efficiency problem of suffix array sorting in the BsDiff algorithm, a fusion algorithm based on the suffix array induced sorting and BsDiff methods was proposed to reduce the time consumption in the OTA upgrade process. The execution of the difference algorithm in the cloud was simulated using bench experiments. Subsequently, the function and performance of the proposed algorithm were experimentally evaluated. The results showed that the optimized algorithm could save more than 35% of the time compared to the original algorithm, which improved the OTA upgrade efficiency to a certain extent.

Abstract Single image dehazing, as a key prerequisite of high-level computer vision tasks, catches more and more attentions. Traditional model-based methods recover haze-free images via atmospheric scattering model, which achieve favorable dehazing effect but endure artifacts, halos, and color distortion. By contrast, recent learning-based methods dehaze images by a model-free way, which achieve better color fidelity but tend to acquire under-dehazed results due to lacking of knowledge guiding. To combine these merits, we propose a novel online knowledge distillation network for single image dehazing named OKDNet. Specifically, the proposed OKDNet firstly preprocesses hazy images and acquires abundant shared features by a multiscale network constructed with attention guided residual dense blocks. After that, these features are sent to different branches to generate two preliminary dehazed images via supervision training: one branch acquires dehazed images via the atmospheric scattering model; another branch directly establishes the mapping relationship between hazy images and clear images, which dehazes images by a model-free way. To effectively fuse useful information from these two branches and acquire a better dehazed results, we propose an efficient feature aggregation block consisted of multiple parallel convolutions with different receptive. Moreover, we adopt a one-stage knowledge distillation strategy named online knowledge distillation to joint optimization of our OKDNet. The proposed OKDNet achieves superior performance compared with state-of-the-art methods on both synthetic and real-world images with fewer model parameters. Project website: https://github.com/lanyunwei/OKDNet .

AimsAbnormal vessel patterns are specific signs in patients with early cervical abnormality and cervical cancer(CC) by colposcopy, but the impact of human papillomavirus (HPV) infections on abnormal vessel patterns remains unknown.MethodsA total of 6716 female patients with HPV infections or cytological abnormalities who underwent a colposcopy following abnormal CC screening results were included in the study. The final pathological diagnosis was confirmed to be the most severe pathological grade across cervical biopsy, endocervical canal curettage (ECC) and conization. Univariate and multivariate logistic regression analyses were used to investigate the association between HPV infections and abnormal vessel patterns, adjusting for age, gravidity and parity.ResultsThere were 6124 normal vascular cases by colposcopy and 592 cases with cervical vascular abnormality. The prevalence of HPV infections was 4284 (70%) in normal patients, and the prevalence of HPV infections was 479 (80%) in cervical vascular abnormality patients. HPV high-risk type 16 infection alone increased the risk of cervical heteromorphic blood vessels (aOR=3.66, 95%CI: 2.54~5.27). HPV 16 and 33 alone (other than the commonly recognized subtype of 18) or coinfection of these two genotypes could increase the risk of cervical punctate vascular and cervical vascular mosaic features and abnormal cervical blood vessels. An increased risk of abnormal cervical lesions was observed for HPV 16 and 33 alone or combined in coinfection compared to the negative group. The risk of cervical vascular abnormality was increased 10-fold by coinfection with HPV 16 and 33 (aOR=10.67, 95% CI: 4.54~25.09, P

Abstract This paper presents a fuzzy fixed‐time learning path tracking control strategy for autonomous vehicles with saturated input and uncertainties. First, Takagi‐Sugeno (T‐S) model is designed to realize an accurate description of the nonlinear path tracking systems. The T‐S model is the weighted sum of a series of linear dynamics models. The weighting coefficients are described by fuzzy membership functions. Then, with the consideration of the saturated input, a fixed‐time learning control is proposed. It contains a learning law to estimate the uncertainties of the system on‐line. The fixed‐time convergence and the robustness properties are analyzed through Lyapunov stability theory. Finally, the effectiveness of the proposed control method has been verified through the CarSim/Simulink co‐simulations under different driving conditions. The results demonstrate that the developed controller significantly improves the performance of path tracking and follows the desired path well even in the case of parameter uncertainties and external disturbances.

Abstract Single‐image dehazing is an important problem because it is a key prerequisite for most high‐level computer vision tasks. Traditional prior‐based methods adopt priors generated from clear images to restrain the atmospheric scattering model and then recover haze‐free images. However, these prior‐based methods always encounter over‐enhancement, such as halos and colour distortion. To solve this problem, many works use a convolutional neural network to retrieve original images. However, without priors as guidance, these learning‐based methods dehaze effectively in synthetic datasets but perform poorly in real scenes. Hence, in this paper, we propose a prior‐guided multiscale network for single‐image dehazing named PGMNet. Specifically, prior‐based methods are adopted to acquire dehazed images of the training dataset in advance and then send these dehazed images to a parameter‐shared encoder to form multiscale features. During the decoding process, these multiscale features are adopted to guide the prior‐guided multiscale network to recover more image details. Moreover, considering that these prior‐based dehazed images usually contain some over‐enhanced regions, a spatial attention guided feature aggregation module and squeeze‐and‐excitation module are adopted to alleviate colour distortion. The proposed PGMNet takes the advantage of prior‐based methods in real haze removal and provides superior performance compared with the state‐of‐the‐art methods on both synthetic and real‐world datasets.

MotivationImage dehazing, as a key prerequisite of high-level computer vision tasks, has gained extensive attention in recent years. Traditional model-based methods acquire dehazed images via the atmospheric scattering model, which dehazed favorably but often causes artifacts due to the error of parameter estimation. By contrast, recent model-free methods directly restore dehazed images by building an end-to-end network, which achieves better color fidelity. To improve the dehazing effect, we combine the complementary merits of these two categories and propose a physical-model guided self-distillation network for single image dehazing named PMGSDN.Proposed methodFirst, we propose a novel attention guided feature extraction block (AGFEB) and build a deep feature extraction network by it. Second, we propose three early-exit branches and embed the dark channel prior information to the network to merge the merits of model-based methods and model-free methods, and then we adopt self-distillation to transfer the features from the deeper layers (perform as teacher) to shallow early-exit branches (perform as student) to improve the dehazing effect.ResultsFor I-HAZE and O-HAZE datasets, better than the other methods, the proposed method achieves the best values of PSNR and SSIM being 17.41dB, 0.813, 18.48dB, and 0.802. Moreover, for real-world images, the proposed method also obtains high quality dehazed results.ConclusionExperimental results on both synthetic and real-world images demonstrate that the proposed PMGSDN can effectively dehaze images, resulting in dehazed results with clear textures and good color fidelity.

Objective: The aim of this study was to discuss the clinical characteristics and the prognosis of surgically diagnosed endometriosis complicated by endometrial polyps and investigate the association between pregnancy outcomes during subsequent pregnancies. Materials and methods: From January 2013 to December 2016, 1263 infertile patients were enrolled in the study. We identified 451 patients with endometriosis, and divided them into a polyp group (n = 204) and a non-polyp group (n = 247) based on whether or not they were associated with endometrial polyps. Postoperative clinical pregnant women (n = 82) among the polyp group were then classified into a study group and a control group composed of those undergoing a singleton pregnancy (n = 164) who delivered during the same time period. Clinical statuses and complications during pregnancy and delivery were collected from hospitals and by telephone interviews and surveys through the mail. Results: The prevalence rate of endometriosis infertile group was obviously higher than the non-endometriosis infertile group ([45.23%; 204/451] versus [17.12%; 139/812]). Women suffering from stage 1 to 4 endometriosis had a 42.44% (73/172), 40.69% (59/145), 55.89% (38/68) and 51.52% (34/66) occurrence rate of endometrial polyps, respectively. The frequency of endometrial polyps for stage 3 and 4 patients was obviously higher than that of stage 1 and 2 patients ([53.73%; 72/134] versus [41.64%; 132/317]). Moreover, the occurrence rate of deep infiltrating endometriosis (DIE) was 57.81% (37/64), which was obviously higher than that of ovarian endometriosis (42.42%; 98/231) and peritoneal endometriosis (44.23%; 69/156). Of the 204 women diagnosed with posterior endometrial polyps, 89 became pregnant, 7 pregnancies ended in a spontaneous abortion, and 82 successfully delivered a baby. The clinical pregnancy rate of patients in stages 1 and 2 was wholly higher than that of patients in stages 3 and 4 ([48.70%; 56/115] versus [37.71%; 26/82]). The postsurgical pregnancy status of patients suffering from peritoneal endometriosis was slightly better than those with ovarian or DIE, but differences were not statistically significant (P = 0.626). We also found that the pregnancy rate was statistically elevated in patients whose EFI scores range from 7 to 10. When compared to the control group, women with endometriosis and endometrial polyps had a higher risk of their pregnancy being complicated by placenta previa (13.41%) and cesarean delivery (59.76%). Conclusion: Patients with endometriosis have a higher frequency of endometrial polyps. We found that a combined hysteroscopy and laparoscopy surgical procedure is an effective way to increase pregnancy rates. Different endometriosis stages and types in patients were associated with clinical pregnancy and spontaneous abortion rates. Women affected by both endometriosis and endometrial polyps have an independently elevated risk of placenta previa and cesarean delivery during pregnancy.

This study is aimed at revealing the dynamic response characteristics of a power-split hybrid electric vehicle (HEV) during the mode transition process with the engine starting. In this paper, for a hybrid powertrain system whose engine is directly connected to a compound power-split transmission, the equivalent lever method is first used to analyze the mode transition process with the engine starting. Then, based on the theoretical formulations and lookup tables, a dynamic model of the compound power-split HEV is established in detail, taking into consideration the engine ripple torque, battery-motor characteristics, dynamic meshing force between gears, and shaft spring-damping characteristics. The powertrain dynamic model is verified by a bench test. Moreover, taking the vehicle jerk and engine starting time as the evaluation indexes of the dynamic response characteristics, the influencing factors of the dynamic response characteristics in the engine starting process are analyzed comprehensively from the aspects of the engine, battery-motor, and engine starting condition. Finally, the engine starting process is synthetically optimized based on the obtained impact law of different influencing factors. The simulation results show that the synthesis optimization method can effectively reduce vehicle jerks and the engine starting time and improve driving comfort during the engine starting process.

Single image dehazing has become a key prerequisite for most high-level computer vision tasks since haze severely degrades the input images. Traditional prior-based methods dehaze images by some assumptions concluded from haze-free images, which recover high-quality results but always cause some halos or color distortion. Recently, many methods have been using convolutional neural networks to learn the haze-relevant features and then retrieve the original images. These learning-based methods achieve better performance in synthetic scenes but can hardly restore a clear image with discriminative texture when applied to real-world images, mainly because these networks are trained on synthetic datasets. To solve these problems, a self-modulated generative adversarial network for single image dehazing named SMGAN is proposed. The SMGAN inputs prior-dehazed images into a parameter-shared encoder to produce some latent information of these dehazed images. During the hazy image decoding process, the latent information is sent to self-modulated batch normalization layers, which makes the network fit in real haze removal. Moreover, consider that there are some over-enhanced regions in the guidance images, and a refine module is proposed to alleviate the negative information. The proposed SMGAN combines the advantages of prior-based methods and learning-based methods, which provides superior performance compared with the state-of-the-art methods on both synthetic and real-word datasets.

The fuel-saving advantages of plug-in hybrid electric vehicles can be improved using optimized configurations and appropriate energy management strategies. However, the noticeable jerks and vibrations of the powertrain can be generated by torque fluctuations in the mode transition (MT), especially, the transition from electric mode to hybrid driving mode that involves engine starting. To address this problem, this paper proposes a dynamic coordinated control strategy that synergizes real-time discrete motor torque change rate limitation (TCRL) and active damping feedback compensation (ADFC) control. First, a detailed vehicle powertrain simulation model is established and validated. The relevant problems in the MT are analyzed by the experimental data. Second, the algorithms relevant to the torque distribution of the power source and the real-time discrete motor TCRL are designed from the kinematic and dynamic relationships of the powertrain at each stage. Considering model inaccuracies, system parameter uncertainties, and load changes, an ADFC is designed based on real-time robust drive shaft torque observer. Moreover, the optimal observer gain is obtained by genetic algorithm under the linear matrix inequalities (LMIs) restriction to improve the robustness of the observer. Finally, the simulation and experimental results indicate that the proposed TCRL-ADFC method can effectively reduce the powertrain shocks and improve the ride comfort.

Objective This study aimed to evaluate the reproductive outcomes of patients who underwent hysteroscopic metroplasty for correction of a complete septate uterus. Methods The study population comprised 92 women with complete septate uteri. Hysteroscopic metroplasty and laparoscopy were performed simultaneously in these patients. The postoperative reproductive outcome of each patient was evaluated. Results In the primary infertility group, there were 32 (40%) pregnancies. In the abortion group, the number of miscarriages decreased from 68 (94.44%) to 5 (10.42%), while the number of live births increased from 1 (1.39%) to 42 (87.50%) after resection compared with before resection. The cumulative probability of pregnancy and that of live-birth pregnancy in the abortion group were significantly higher than those in the primary infertility group after surgery. Furthermore, resection of the cervical septum resulted in a significantly higher cumulative probability of live birth compared with preservation of the cervical septum. Conclusion Hysteroscopic uterine metroplasty may improve the reproductive performance of a septate uterus. Resection of the cervical septum may increase the probability of a live-birth pregnancy for patients with a cervical septum, and this procedure could be recommended for cases of a complete uterine septum.

Experimental research is essential in the development of a hybrid electric vehicle. In this study, a bench test was conducted for a compound power-split hybrid electric vehicle (PSHEV) to analyze the real dynamic characteristics of its components and the factors of system shock and vibration during the engine start-up process. Firstly, the mode switching process with an engine start-up was divided into four stages by the lever method. The basic control strategy of mode switching with engine start-up was formulated and tested on a bench test platform. Secondly, based on the bench test data, the output characteristics of the battery motor, engine, and driveshaft were analyzed in detail. The main variable parameters of the engine control unit were investigated in the engine start-up process. Ultimately, the results showed that the engine’s pulsating torque was the main reason for system jerk and vibration during the engine start-up process, and the excessive intake manifold pressure before the engine’s ignition was one of the main reasons for the large output torque ripple. When initiating the electric engine starting process, the jerk and vibration presented a wide fluctuation. The maximum value of the equivalent jerk was 92.12 m/s3, and the maximum value of the absolute value of the vibration acceleration was 4.077 m/s2.

For a power-split plug-in hybrid electric vehicle (PS-PHEV), the control system cannot effectively reduce jerks under some engine-starting conditions due to the pulsating torque of the engine, elastic characteristics of the transmission, battery power limitations, and the limitation of the maximum torque of the motor. To solve this problem, a dynamic torque-coordinated control method considering engine starting conditions (DTCC-ESC) is proposed in this paper. The proposed DTCC-ESC mainly includes three parts: the engine segment active control before and after engine ignition to minimize the impact of the engine ripple torque on the shocks of the powertrain system; the feedforward control of the engine starting conditions and the feedback control of the engine’s optimal target speed; and the active damping feedback compensation control for system resonance to reduce system vibrations. The results show that the proposed DTCC-ESC can effectively improve driving comfort under different engine starting conditions.

The revolution of physical structure is highly significant for future software defined vehicles (SDV). Active structural transformation is a promising feature of the next generation of vehicle physical structure. It can enhance the dynamic performance of vehicles, thus providing safer and more comfortable ride experiences, such as the ability to avoid rollover in critical situations. Based on the active structural transformation technology, this study proposes a novel approach to improve the dynamic performance of a vehicle. The first analytical motion model of a vehicle with active structural transformation capability is established. Then, a multi-objective optimization problem with the adjustable parameters as design variables is abstracted and solved with an innovative scenario specific optimization method. Simulation results under different driving scenarios revealed that the active transformable vehicle applying the proposed method could significantly improve the handling stability without sacrificing the ride comfort, compared with a conventional vehicle with a fixed structure. The proposed method pipeline is defined by the software and supported by the hardware. It fully embodies the characteristics of SDV, and inspires the improvement of multiple types of vehicle performance based on the concept of "being defined by software" and the revolution of the physical structure. [ABSTRACT FROM AUTHOR]