Search

Abstract To achieve high power conversion efficiency in perovskite/silicon tandem solar cells, it is necessary to develop a promising wide-bandgap perovskite absorber and processing techniques in relevance. To date, the performance of devices based on wide-bandgap perovskite is still limited mainly by carrier recombination at their electron extraction interface. Here, we demonstrate assembling a binary two-dimensional perovskite by both alternating-cation-interlayer phase and Ruddlesden−Popper phase to passivate perovskite/C60 interface. The binary two-dimensional strategy takes effects not only at the interface but also in the bulk, which enables efficient charge transport in a wide-bandgap perovskite solar cell with a stabilized efficiency of 20.79% (1 cm2). Based on this absorber, a monolithic perovskite/silicon tandem solar cell is fabricated with a steady-state efficiency of 30.65% assessed by a third party. Moreover, the tandem devices retain 96% of their initial efficiency after 527 h of operation under full spectral continuous illumination, and 98% after 1000 h of damp-heat testing (85 °C with 85% relative humidity).

The expansion of the digital economy is crucial for halting climate change, as carbon emissions from urban energy use contribute significantly to global warming. This study uses the Difference-in-Differences Model and the Spatial Durbin Model determine whether the digital economy may support the development of reducing carbon emissions and its geographic spillover effects in Chinese cities on the east coast. In addition, it looks more closely at the effects of lowering carbon emissions in space by separating them into direct, indirect, and spatial impact parts. The findings show that (1) from 2012 to 2021, the digital economy favored carbon emission reductions in China’s eastern coastline cities, as supported by the robustness test. (2) The link between digital economy growth and carbon emissions is highly variable, with smart city development and urban agglomeration expansion both cutting city carbon emissions considerably. Successful digital economy strategies can lower CO2 emissions from nearby cities. (3) Eastern coastal cities have a considerable spatial spillover impact, and the digital economy mitigates local energy consumption and carbon emissions while simultaneously enhancing environmental quality in nearby urban areas. This analysis proposes that the peak carbon and carbon neutrality targets can be met by increasing the digital economy and enhancing regional environmental governance cooperation.

The Cambrian subsalt dolomite in the Kalpin area of the Tarim Basin is an important reserve growth point and strategic replacement area. However, there is a lack of clear understanding regarding the formation mechanism of high-quality reservoirs in this region, which has hindered oil and gas exploration. This study aims to address this knowledge gap by providing a comprehensive description of the rock types and characteristics of the Xiao'erbulak Formation. Microscope observations, geochemical analyses, and interpretations of well logging data from Well KPN1 were used in this analysis. The Xiao'erbulak Formation can be divided into four members, arranged from bottom to top. The relatively high manganese (Mn) content (87–137.7 ppm), oxygen isotope composition (δ18O) with an average value of −6.37‰, and strontium isotope ratio (87Sr/86Sr) with an average value of 0.7109 indicate that dolomitization of the Xiao'erbulak Formation likely occurred during the penecontemporaneous-shallow burial period. The early formation of dolomite contributed to increased reservoir porosity and resistance to compaction during deep burial, which laid the foundation for reservoir formation. The carbon isotope composition (δ13C) of the Xiao 1 and Xiao 2 members exhibit frequent zigzag curves, indicating recurring progression/regression processes. The subsequent development of granular beach facies played a crucial role in reservoir formation in the Xiao 2 Member. Tectonic fractures and penecontemporaneous karstification controlled the reservoir characteristics of the Xiao 3 Member. Furthermore, this study provides an analysis of the diagenetic evolution model of Well KPN1 and examines the impact of diagenetic transformations on reservoir quality. The systematic analysis of downhole data from Well KPN1 serves as a foundational reference for comparative studies with other drilling sites in the area. It also offers valuable guidance for future exploration and deployment strategies in the northwest Tarim Basin.

Abstract Data scarcity is a major challenge when training deep learning (DL) models. DL demands a large amount of data to achieve exceptional performance. Unfortunately, many applications have small or inadequate data to train DL frameworks. Usually, manual labeling is needed to provide labeled data, which typically involves human annotators with a vast background of knowledge. This annotation process is costly, time-consuming, and error-prone. Usually, every DL framework is fed by a significant amount of labeled data to automatically learn representations. Ultimately, a larger amount of data would generate a better DL model and its performance is also application dependent. This issue is the main barrier for many applications dismissing the use of DL. Having sufficient data is the first step toward any successful and trustworthy DL application. This paper presents a holistic survey on state-of-the-art techniques to deal with training DL models to overcome three challenges including small, imbalanced datasets, and lack of generalization. This survey starts by listing the learning techniques. Next, the types of DL architectures are introduced. After that, state-of-the-art solutions to address the issue of lack of training data are listed, such as Transfer Learning (TL), Self-Supervised Learning (SSL), Generative Adversarial Networks (GANs), Model Architecture (MA), Physics-Informed Neural Network (PINN), and Deep Synthetic Minority Oversampling Technique (DeepSMOTE). Then, these solutions were followed by some related tips about data acquisition needed prior to training purposes, as well as recommendations for ensuring the trustworthiness of the training dataset. The survey ends with a list of applications that suffer from data scarcity, several alternatives are proposed in order to generate more data in each application including Electromagnetic Imaging (EMI), Civil Structural Health Monitoring, Medical imaging, Meteorology, Wireless Communications, Fluid Mechanics, Microelectromechanical system, and Cybersecurity. To the best of the authors’ knowledge, this is the first review that offers a comprehensive overview on strategies to tackle data scarcity in DL.

The spatial auto-correlation analysis approach is used to examine the agglomeration features and geographical correlation of industrial agglomeration and pollutant discharge using 91 prefecture-level cities in the Yellow River Basin from 2005 to 2020 as the research target. The Yellow River Basin and the upper, middle, and lower reaches were investigated for the effects of industrial agglomeration on industrial pollution emissions. The Dubin model of the spatial panel is used to investigate the drivers of industrial pollution from both indirect effects and direct effects. The findings indicated that ① the spatial pattern of industrial pollution agglomeration and industrial agglomeration in the Yellow River Basin has spatial convergence; ② the global clustering characteristics of industrial pollution in the Yellow River Basin are apparent. The global correlations of the three pollutants are industrial wastewater > industrial smoke and dust > industrial SO2; there is a spatial association between industrial agglomeration and pollution agglomeration, and there are differences in the spatial interconnection between industrial agglomeration and the accumulation of different pollutants; ③ from the overall point of view of the Yellow River Basin, industrial agglomeration will reduce the pollution levels of industrial wastewater, industrial SO2, and industrial smoke and dust. From the perspective of different regions, industrial agglomeration in the upper, middle, and lower reaches is conducive to the decline of industrial pollution, and the role is in the middle, downstream, and upstream in order from large to small; and ④ the Yellow River Basin’s industrial pollution is primarily caused by population agglomeration, industrial structure, environmental regulation, opening to the outside world, and economic development. Their effects on industrial wastewater, industrial SO2, industrial smoke, and dust pollution vary.

While the research on industrial agglomeration has received more attention, there has been little research on its spatial effect and impact mechanism from the micro perspective. On the basis of the industrial cluster data of 600 counties in the Yellow River basin (YRB) from 2010 to 2020, this article investigates the spatial effect of industrial agglomeration in the YRB from the two aspects of time and space via center of gravity analysis and exploratory spatial data analysis and uses the spatial econometric method to research the drivers of county-level industrial agglomeration. The results reveal the following. During the study period, the industrial center of gravity in the YRB presented a moving tendency from east to west, and 2016 was an important turning point. The county-level industrial distribution in the YRB showed an unbalanced trend, but its unbalanced degree exhibited a weakening trend. In terms of the temporal dimension, the county-level industrial agglomeration degree in the YRB presented a significant downward trend and emerged an evolution process of “diffusion–agglomeration–re-diffusion.” Specifically, it gradually diffused from Shandong and Henan to southern Shanxi and Shaanxi. In the spatial dimension, the county-level industrial agglomeration in the YRB was found to have conspicuous spatial autocorrelation, and the spatial spillover effect was prominent. The number of counties with high–high cluster gradually increased and shifted to the west, while the number of counties with low–low agglomeration exhibited a decreasing trend. The degree of government intervention restrains county-level industrial agglomeration, while the external industry level, regional market demand, and urbanization level promote county-level industrial agglomeration. In terms of river basins, in the upstream, the degree of government intervention and industrial structure adjustment inhibit industrial agglomeration, while the regional market demand and urbanization level promote industrial agglomeration. In the midstream, the degree of government intervention inhibits industrial agglomeration, while the external industry level, regional market demand, and industrial structure adjustment promote industrial agglomeration. In the downstream, the external industry level and industrial structure adjustment restrain industrial agglomeration, while the regional market demand and urbanization level promote industrial agglomeration. The conclusions provide a significant reference for the industrial transformation and high-quality economic development of the YRB.

Abstract In the last few years, the deep learning (DL) computing paradigm has been deemed the Gold Standard in the machine learning (ML) community. Moreover, it has gradually become the most widely used computational approach in the field of ML, thus achieving outstanding results on several complex cognitive tasks, matching or even beating those provided by human performance. One of the benefits of DL is the ability to learn massive amounts of data. The DL field has grown fast in the last few years and it has been extensively used to successfully address a wide range of traditional applications. More importantly, DL has outperformed well-known ML techniques in many domains, e.g., cybersecurity, natural language processing, bioinformatics, robotics and control, and medical information processing, among many others. Despite it has been contributed several works reviewing the State-of-the-Art on DL, all of them only tackled one aspect of the DL, which leads to an overall lack of knowledge about it. Therefore, in this contribution, we propose using a more holistic approach in order to provide a more suitable starting point from which to develop a full understanding of DL. Specifically, this review attempts to provide a more comprehensive survey of the most important aspects of DL and including those enhancements recently added to the field. In particular, this paper outlines the importance of DL, presents the types of DL techniques and networks. It then presents convolutional neural networks (CNNs) which the most utilized DL network type and describes the development of CNNs architectures together with their main features, e.g., starting with the AlexNet network and closing with the High-Resolution network (HR.Net). Finally, we further present the challenges and suggested solutions to help researchers understand the existing research gaps. It is followed by a list of the major DL applications. Computational tools including FPGA, GPU, and CPU are summarized along with a description of their influence on DL. The paper ends with the evolution matrix, benchmark datasets, and summary and conclusion.

Abstract Background Cryo-EM data generated by electron tomography (ET) contains images for individual protein particles in different orientations and tilted angles. Individual cryo-EM particles can be aligned to reconstruct a 3D density map of a protein structure. However, low contrast and high noise in particle images make it challenging to build 3D density maps at intermediate to high resolution (1–3 Å). To overcome this problem, we propose a fully automated cryo-EM 3D density map reconstruction approach based on deep learning particle picking. Results A perfect 2D particle mask is fully automatically generated for every single particle. Then, it uses a computer vision image alignment algorithm (image registration) to fully automatically align the particle masks. It calculates the difference of the particle image orientation angles to align the original particle image. Finally, it reconstructs a localized 3D density map between every two single-particle images that have the largest number of corresponding features. The localized 3D density maps are then averaged to reconstruct a final 3D density map. The constructed 3D density map results illustrate the potential to determine the structures of the molecules using a few samples of good particles. Also, using the localized particle samples (with no background) to generate the localized 3D density maps can improve the process of the resolution evaluation in experimental maps of cryo-EM. Tested on two widely used datasets, Auto3DCryoMap is able to reconstruct good 3D density maps using only a few thousand protein particle images, which is much smaller than hundreds of thousands of particles required by the existing methods. Conclusions We design a fully automated approach for cryo-EM 3D density maps reconstruction (Auto3DCryoMap). Instead of increasing the signal-to-noise ratio by using 2D class averaging, our approach uses 2D particle masks to produce locally aligned particle images. Auto3DCryoMap is able to accurately align structural particle shapes. Also, it is able to construct a decent 3D density map from only a few thousand aligned particle images while the existing tools require hundreds of thousands of particle images. Finally, by using the pre-processed particle images, Auto3DCryoMap reconstructs a better 3D density map than using the original particle images.

Abstract Background Cryo-electron microscopy (Cryo-EM) is widely used in the determination of the three-dimensional (3D) structures of macromolecules. Particle picking from 2D micrographs remains a challenging early step in the Cryo-EM pipeline due to the diversity of particle shapes and the extremely low signal-to-noise ratio of micrographs. Because of these issues, significant human intervention is often required to generate a high-quality set of particles for input to the downstream structure determination steps. Results Here we propose a fully automated approach (DeepCryoPicker) for single particle picking based on deep learning. It first uses automated unsupervised learning to generate particle training datasets. Then it trains a deep neural network to classify particles automatically. Results indicate that the DeepCryoPicker compares favorably with semi-automated methods such as DeepEM, DeepPicker, and RELION, with the significant advantage of not requiring human intervention. Conclusions Our framework combing supervised deep learning classification with automated un-supervised clustering for generating training data provides an effective approach to pick particles in cryo-EM images automatically and accurately.

In this paper, we introduce REDN: A Recursive Encoder-Decoder Network with Skip-Connections for edge detection in natural images. The proposed network is a novel integration of a Recursive Neural Network with an Encoder-Decoder architecture. The recursive network enables iterative refinement of the edges using a single network model. Adding skip-connections between encoder and decoder helps the gradients reach all the layers of a network more easily and allows information related to finer details in the early stage of the encoder to be fully utilized in the decoder. Based on our extensive experiments on popular boundary detection datasets including BSDS500 [1], NYUD [2] and Pascal Context [3], REDN significantly advances the state-of-the-art on edge detection regarding standard evaluation metrics such as Optimal Dataset Scale (ODS) F-measure, Optimal Image Scale (OIS) F-measure, and Average Precision (AP).

Based on the data of existing warehouse supermarkets in Liaoning Province, China, spatial autocorrelation analysis, kernel density analysis, composite correlation coefficient analysis and other methods have been adopted to analyze their spatial-temporal evolution pattern to reflect the general law of the development of China’s existing warehouse supermarkets and fill the gap in this research field. The results show that the spatial distribution of warehouse supermarkets in Liaoning Province is extremely uneven, and areas with high nuclear density are distributed along the “Shenyang-Dalian” line belonging to the aggregation distribution. The Lorentz curve shows a downward trend with a large degree of spatial imbalance, that is, the regional concentration of warehouse supermarkets is high. Through global and local autocorrelation analysis, the regions with similar development levels of warehouse supermarkets in Liaoning Province tend to gather together, and the spatial distribution has a strong correlation. The distribution of warehousing supermarkets in Liaoning Province is affected by traffic location conditions, economic conditions, population quantity and population density, the number of urban functional areas, policy conditions and the role of the government, especially by economic conditions.

Transfer learning (TL) has been widely utilized to address the lack of training data for deep learning models. Specifically, one of the most popular uses of TL has been for the pre-trained models of the ImageNet dataset. Nevertheless, although these pre-trained models have shown an effective performance in several domains of application, those models may not offer significant benefits in all instances when dealing with medical imaging scenarios. Such models were designed to classify a thousand classes of natural images. There are fundamental differences between these models and those dealing with medical imaging tasks regarding learned features. Most medical imaging applications range from two to ten different classes, where we suspect that it would not be necessary to employ deeper learning models. This paper investigates such a hypothesis and develops an experimental study to examine the corresponding conclusions about this issue. The lightweight convolutional neural network (CNN) model and the pre-trained models have been evaluated using three different medical imaging datasets. We have trained the lightweight CNN model and the pre-trained models with two scenarios which are with a small number of images once and a large number of images once again. Surprisingly, it has been found that the lightweight model trained from scratch achieved a more competitive performance when compared to the pre-trained model. More importantly, the lightweight CNN model can be successfully trained and tested using basic computational tools and provide high-quality results, specifically when using medical imaging datasets.

China’s steel industry has not yet implemented a carbon tax policy, and its benefits and impacts are still in the theoretical research stage. In addition, enterprises have an insufficient ability to respond to changes in production and sales, which seriously affects the market’s stability. The government should simultaneously start from multiple perspectives, such as energy conservation, emission reduction, dynamic adjustments, and business decisions. Therefore, this research constructs a repeated dynamic game model including carbon tax policy and other mixed reduction policies, and studies the stability and related indicators of the market. The results are as follows: (1) the output adjustment policies that enterprises can implement will show an increasing trend under the single carbon tax policy. (2) The output adjustment policies that enterprises with larger output will also show an increasing trend under the mixed emission reduction policy when emission reduction targets continue to increase. (3) Smaller-output enterprises need to be more cautious in formulating their production plans, and their output adjustment policies will be restricted and affected by more factors. In summary, enterprises should comprehensively consider emission reduction policies, output adjustment policies and other enterprises’ output changes, to ensure that the steel market will not fall into an imbalanced state.

With China’s carbon peak and target proposed and the power system as an important source of carbon emissions, its carbon reduction issues are of great concern. However, the mismatch between power demand, resources, and generation facilities in different regions of China creates enormous carbon reduction pressures, especially in the region covered by China’s Northeast Power Grid (NE Grid). Therefore, in this study, a two-stage power optimization model considering Carbon Capture, Utilization, and Storage (CCUS) was established to optimize the national power generation and transmission in China from 2020 to 2060 first and then to optimize the regional road map of energy structure adjustment for China’s NE grid and its CO2 storage strategy. The first stage optimization results show that the national power interaction in China can make full use of the energy resource advantages of the NE grid, especially Jilin and Mengdong, delivering clean electricity to the North China grid as it reaches its “dual carbon” goal. Additionally, the second-stage optimization results show that the NE grid can make good use of CCUS for its thermal power transition. Annual CO2 storage will peak at 204 Mt in 2050 as thermal power withdraws in the NE grid region, which provides a reference for other regions to realize carbon neutrality of their power system by using CCUS.

At present, the problems of homogenization and low quality in China’s iron and steel industry are particularly prominent and the ability of the enterprises to cope with change is insufficient. Adopting product differentiation strategy and dynamic adjustment strategy can allow steel enterprises and the industry to better adapt to future changes. By introducing the product differentiation degree (substitution coefficient) and the bounded rationality strategy to simulate these two strategic means, this paper constructs an extended two-stage dynamic game model to analyse the dynamic game scenarios and steel market stability in China. As new findings, we report the following: (1) The system is more likely to fall into an unbalanced state when multiple enterprises adopt the policy of dynamic output adjustment simultaneously. (2) Enterprises with large output and small output have different output adjustment policies. When enterprises with big-scale output adopt a bit larger adjustment policies, enterprises with small output will be strongly impacted, and the available adjustment space will be sharply compressed. (3) The gradual increase in the difference between products reduces the stability of the market. (4) When product differentiation and bounded rationality strategies coexist, the steel market may fall into an unbalanced state when the degree of product difference increases excessively and the enterprise adopts more drastic output adjustment policies. Therefore, there are pros and cons to product differentiation strategy and bounded rationality adjustment strategy. When each steel oligopoly enterprise formulates a production plan, it needs to comprehensively consider the output changes of the other enterprises and carefully weigh the strategic issues.

Rural tourism is a new point of growth for tourism and the economy in the context of the new normalization of the economy and is of great significance in achieving the complementary coordination and integration of urban and rural areas, promoting rural transformation, and increasing farmers’ incomes. The trends of rural tourism development mechanisms studied on a spatial scale can be used to interpret the sustainable development of rural tourism from different perspectives. Based on the data of key rural tourism villages in China’s Yellow River Basin (hereinafter referred to as the Yellow River Basin), kernel density estimation and spatial hot spot clustering methods were used in the present study to analyze the spatial distribution pattern and sustainable development mechanisms of these villages. The results showed that the spatial distribution of the key villages presents greater concentrations in the west and south than in the east and north, respectively. The spatial distribution of the key villages was found to be primarily affected by factors such as historical culture, transportation locations, economic level, and topography. Finally, the sustainable development mechanisms of rural tourism are proposed, and corresponding suggestions are provided from the perspectives of sustainable livelihoods, operation management, and marketing.

One of the main challenges of employing deep learning models in the field of medicine is a lack of training data due to difficulty in collecting and labeling data, which needs to be performed by experts. To overcome this drawback, transfer learning (TL) has been utilized to solve several medical imaging tasks using pre-trained state-of-the-art models from the ImageNet dataset. However, there are primary divergences in data features, sizes, and task characteristics between the natural image classification and the targeted medical imaging tasks. Therefore, TL can slightly improve performance if the source domain is completely different from the target domain. In this paper, we explore the benefit of TL from the same and different domains of the target tasks. To do so, we designed a deep convolutional neural network (DCNN) model that integrates three ideas including traditional and parallel convolutional layers and residual connections along with global average pooling. We trained the proposed model against several scenarios. We utilized the same and different domain TL with the diabetic foot ulcer (DFU) classification task and with the animal classification task. We have empirically shown that the source of TL from the same domain can significantly improve the performance considering a reduced number of images in the same domain of the target dataset. The proposed model with the DFU dataset achieved F1-score value of 86.6% when trained from scratch, 89.4% with TL from a different domain of the targeted dataset, and 97.6% with TL from the same domain of the targeted dataset.

To study the emission reduction policies’ impact on the production and economic level of the steel industry, this paper constructs a two-stage dynamic game model and analyzes various emission reduction policies’ impact on the steel industry and enterprises. New results are observed in the study: (1) With the increasing emission reduction target (15%−30%) and carbon quota trading price (12.65−137.59 Yuan), social welfare and producer surplus show an increasing trend and emission macro losses show a decreasing trend. (2) Enterprises’ reduction ranges in northwestern and southwestern regions are much higher than that of the other regions; the northeastern enterprise has the smallest reductions range. (3) When the market is balanced (0.8543−0.9320 billion tons), the steel output has decreased and the polarization in various regions has been alleviated to some extent. The model is the abstraction and assumption of reality, which makes the results have some deviations. However, these will provide references to formulate reasonable emissions reduction and production targets. In addition, the government needs to consider the whole and regional balance and carbon trading benchmark value when deciding the implementation of a single or mixed policy. Future research will be more closely linked to national policies and gradually extended to other high-energy industries.

In this paper, a two-stage dynamic game model of China’s iron and steel industry is constructed. Carbon tax levy, product subsidy, carbon capture and sequestration (CCS) and other factors are included in the emission reduction mechanism. The effects of emissions reduction and the economic impact of China’s overall steel industry (and that of its six main regions) are investigated for the first time under different scenarios. As new findings, we report the following: (1) Not all factors declined. The overall social welfare, consumer surplus, output and emissions decrease with a gradual increase in the reduction target, whereas the carbon tax value, unit value of product subsidies and total subsidies show a rising trend; (2) A combination of multiple emissions reduction policies is more effective than a single policy. With the implementation of a combined policy, regional output polarization has eased; (3) Steel output does not exceed 950 million tons, far below the current peak. These results will help the industry to formulate reasonable emissions reduction and output targets. In short, in effort to eliminate industry poverty and to alleviate overcapacity, the industry should not only adopt the various coordinated reduction policies, but also fully consider regional differences and reduction needs.

The dystrophin-deficient dog is excellent large animal model for testing novel therapeutic modalities for Duchenne muscular dystrophy (DMD). Despite well-documented descriptions of dystrophic symptoms in these dogs, very few quantitative studies have been performed. Here, we developed a comprehensive set of non-invasive assays to quantify dog gait (stride length and speed), joint angle and limb mobility (for both forelimb and hind limb), and spontaneous activity at night. To validate these assays, we examined three 8-m-old mix-breed dystrophic dogs. We also included three age-matched siblings as the normal control. High-resolution video recorders were used to digitize dog walking and spontaneous movement at night. Stride speed and length were significantly decreased in affected dogs. The mobility of the limb segments (forearm, front foot, lower thigh, rear foot) and the carpus and hock joints was significantly reduced in dystrophic dogs. There was also a significant reduction of the movement in affected dogs during overnight monitoring. In summary, we have established a comprehensive set of outcome measures for clinical phenotyping of DMD dogs. These non-invasive end points would be valuable in monitoring disease progression and therapeutic efficacy in translational studies in the DMD dog model.

We propose a new iterative surface evolution algorithm for multiview stereo. Starting from an embedding space such as the visual hull, we will first conduct robust 3D depth estimation (represented as 3D points) based on image correlation. A fast implicit distance function-based region growing method is then employed to extract an initial shape estimation based on these 3D points. Next, an explicit surface evolution will be conducted to recover the finer geometry details of the recovered shape. The recovered shape will be further improved by several iterations between depth estimation and shape reconstruction, similar to the Expectation Maximization (EM) approach. The experiments on the benchmark datasets show that our algorithm can obtain high-quality reconstruction results that are comparable with the state-of-art methods, with considerable less computational time and complexity.

We propose a semi-automatic thalamus and thalamus nuclei segmentation algorithm from Diffusion Tensor Magnetic Resonance Imaging (DT-MRI) based on the mean-shift algorithm. Comparing with existing thalamus segmentation algorithms which are mainly based on K-means algorithm, our mean-shift based algorithm is more flexible and adaptive. It does not assume a Gaussian distribution or a fixed number of clusters. Furthermore, the single parameter in the mean-shift based algorithm supports hierarchical clustering naturally.

We propose a new context-sensitive active contour for 2D corpus callosum segmentation. After a seed contour consisting of interconnected parts is being initialized by the user, each part will start to deform according to its own motion law derived from high-level prior knowledge, and is constantly aware of its own orientation and destination during the deformation process. Experimental results demonstrate the accuracy and robustness of our algorithm.

We present a method for semiautomatic segmentation of brain structures such as thalamus from MRI images based on the concept of geometric surface flow. Given an MRI image, the user can interactively initialize a seed model within region of interest. The model will then start to evolve by incorporating both boundary and region information following the principle of variational analysis. The deformation will stop when an equilibrium state is achieved. To overcome the low contrast of the original image data, a nonparametric kernel-based method is applied to simultaneously update the interior probability distribution during the model evolution. Our experiments on both 2D and 3D image data demonstrate that the new method is robust to image noise and inhomogeneity and will not leak from spurious edge gaps.

We recently developed a multiresolution PDE-based deformable surface whose deformation behavior is governed by partial differential equations (PDEs) such as the weighted minimal surface flow. Comparing with the level-set approach, our new model has better control of the mesh quality and model resolution, and is much simpler to implement since all the computations are local. The new deformable model is very useful for a variety of medical imaging applications including boundary reconstruction, surface visualization, data segmentation, and topology discovery. In this paper, we demonstrate both the accuracy and robustness of our model on areas such as medical image segmentation through a number of experiments on both real (MRI/CT) and synthetic volumetric datasets.