Search

Background: The clinical outcomes of bipolar radiofrequency (RF) lipolysis, a prevalent non-invasive fat reduction procedure, hinge on the delicate balance between effective lipolysis and patient safety, with skin overheating and subsequent tissue damage as primary concerns., Objective: This study aimed to investigate a novel bipolar radiofrequency lipolysis technique, safeguarding the skin through an innovative PID temperature control algorithm., Methods: Utilizing COMSOL Multiphysics simulation software, a two-dimensional fat and skin tissue model was established, simulating various PID temperature control schemes. The crux of the simulation involved a comparative analysis of different PID temperatures at 45 °C, 50 °C, and 55 °C and constant power strategies, assessing their implications on skin temperature. Concurrently, a custom bipolar radiofrequency lipolysis device was developed, with ex vivo experiments conducted using porcine tissue for empirical validation., Results: The findings indicated that with PID settings of Kp = 7, Ki = 2, and Kd = 0, and skin temperature control at 45 °C or 50 °C, the innovative PID-based epidermal temperature control strategy successfully maintained the epidermal temperature within a safe range. This maintenance was achieved without compromising the effectiveness of RF lipolysis, significantly reducing the risk of thermal damage to the skin layers., Conclusion: Our research confirms the substantial practical utility of this advanced PID-based bipolar RF lipolysis technique in clinical aesthetic procedures, enhancing patient safety during adipose tissue ablation therapies.

As a powerful but computationally intensive method, hybrid computational models study the dynamics of multicellular systems by evolving discrete cells in reacting and diffusing extracellular microenvironments. As the scale and complexity of studied biological systems continuously increase, the exploding computational cost starts to limit large-scale cell-based simulations. To facilitate the large-scale hybrid computational simulation and make it feasible on easily accessible computational devices, we develop Gell (GPU Cell), a fast and memory-efficient open-source GPU-based hybrid computational modeling platform for large-scale system modeling. We fully parallelize the simulations on GPU for high computational efficiency and propose a novel voxel sorting method to further accelerate the modeling of massive cell-cell mechanical interaction with negligible additional memory footprint. As a result, Gell efficiently handles simulations involving tens of millions of cells on a personal computer. We compare the performance of Gell with a state-of-the-art paralleled CPU-based simulator on a hanging droplet spheroid growth task and further demonstrate Gell with a ductal carcinoma in situ (DCIS) simulation. Gell affords ~150X acceleration over the paralleled CPU method with one-tenth of the memory requirement., Competing Interests: The authors have declared that no competing interests exist., (Copyright: © 2023 Du et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.)

Leveraging machine learning techniques for Sepsis early detection and diagnosis has attracted increasing interest in recent years. However, most existing methods require a large amount of labeled training data, which may not be available for a target hospital that deploys a new Sepsis detection system. More seriously, as treated patients are diversified between hospitals, directly applying a model trained on other hospitals may not achieve good performance for the target hospital. To address this issue, we propose a novel semi-supervised transfer learning framework based on optimal transport theory and self-paced ensemble for Sepsis early detection, called SPSSOT, which can efficiently transfer knowledge from the source hospital (with rich labeled data) to the target hospital (with scarce labeled data). Specifically, SPSSOT incorporates a new optimal transport-based semi-supervised domain adaptation component that can effectively exploit all the unlabeled data in the target hospital. Moreover, self-paced ensemble is adapted in SPSSOT to alleviate the class imbalance issue during transfer learning. In a nutshell, SPSSOT is an end-to-end transfer learning method that automatically selects suitable samples from two domains (hospitals) respectively and aligns their feature spaces. Extensive experiments on two open clinical datasets, MIMIC-III and Challenge, demonstrate that SPSSOT outperforms state-of-the-art transfer learning methods by improving 1-3% of AUC.

Clustering is a powerful machine learning method for discovering similar patterns according to the proximity of elements in feature space. It is widely used in computer science, bioscience, geoscience, and economics. Although the state-of-the-art partition-based and connectivity-based clustering methods have been developed, weak connectivity and heterogeneous density in data impede their effectiveness. In this work, we propose a boundary-seeking Clustering algorithm using the local Direction Centrality (CDC). It adopts a density-independent metric based on the distribution of K-nearest neighbors (KNNs) to distinguish between internal and boundary points. The boundary points generate enclosed cages to bind the connections of internal points, thereby preventing cross-cluster connections and separating weakly-connected clusters. We demonstrate the validity of CDC by detecting complex structured clusters in challenging synthetic datasets, identifying cell types from single-cell RNA sequencing (scRNA-seq) and mass cytometry (CyTOF) data, recognizing speakers on voice corpuses, and testifying on various types of real-world benchmarks., (© 2022. The Author(s).)

Highlight removal is a fundamental and challenging task that has been an active field for decades. Although several methods have recently been improved for facial images, they are typically designed for a single image. This paper presents a lightweight optimization method for removing the specular highlight reflections of multi-view facial images. This is achieved by taking full advantage of the Lambertian consistency, which states that the diffuse component does not vary with the change in the viewing angle, while the specular component changes the behavior. We provide non-negative constraints on light and shading in all directions, rather than normal directions contained in the face, to obtain physically reliable properties. The removal of highlights is further facilitated through the estimation of illumination chromaticity, which is done by employing orthogonal subspace projection. An important practical feature of the proposed method does not require face reflectance priors. A dataset with ground truth for highlight removal of multi-view facial images is captured to quantitatively evaluate the performance of our method. We demonstrate the robustness and accuracy of our method through comparisons to existing methods for removing specular highlights and improvement in applications such as reconstruction.

Multimodal optimization problems (MMOPs) require algorithms to locate multiple optima simultaneously. When using evolutionary algorithms (EAs) to deal with MMOPs, an intuitive idea is to divide the population into several small "niches," where different niches focus on locating different optima. These population partition strategies are called "niching" techniques, which have been frequently used for MMOPs. The algorithms for simultaneously locating multiple optima of MMOPs are called multimodal algorithms. However, many multimodal algorithms still face the difficulty of population partition since most of the niching techniques involve the sensitive niching parameters. Considering this issue, in this article, we propose a parameter-free niching method based on adaptive estimation distribution (AED) and develop a distributed differential evolution (DDE) algorithm, which is called AED-DDE, for solving MMOPs. In AED-DDE, each individual finds its own appropriate niche size to form a niche and acts as an independent unit to find a global optimum. Therefore, we can avoid the difficulty of population partition and the sensitivity of niching parameters. Different niches are co-evolved by using the master-slave multiniche distributed model. The multiniche co-evolution mechanism can improve the population diversity for fully exploring the search space and finding more global optima. Moreover, the AED-DDE algorithm is further enhanced by a probabilistic local search (PLS) to refine the solution accuracy. Compared with other multimodal algorithms, even the winner of CEC2015 multimodal competition, the comparison results fully demonstrate the superiority of AED-DDE.

Aiming at the problems of a large volume, slow processing speed, and difficult deployment in the edge terminal, this paper proposes a lightweight insulator detection algorithm based on an improved SSD. Firstly, the original feature extraction network VGG-16 is replaced by a lightweight Ghost Module network to initially achieve the lightweight model. A Feature Pyramid structure and Feature Pyramid Network (FPN+PAN) are integrated into the Neck part and a Simplified Spatial Pyramid Pooling Fast (SimSPPF) module is introduced to realize the integration of local features and global features. Secondly, multiple Spatial and Channel Squeeze-and-Excitation (scSE) attention mechanisms are introduced in the Neck part to make the model pay more attention to the channels containing important feature information. The original six detection heads are reduced to four to improve the inference speed of the network. In order to improve the recognition performance of occluded and overlapping targets, DIoU-NMS was used to replace the original non-maximum suppression (NMS). Furthermore, the channel pruning strategy is used to reduce the unimportant weight matrix of the model, and the knowledge distillation strategy is used to fine-adjust the network model after pruning, so as to ensure the detection accuracy. The experimental results show that the parameter number of the proposed model is reduced from 26.15 M to 0.61 M, the computational load is reduced from 118.95 G to 1.49 G, and the mAP is increased from 96.8% to 98%. Compared with other models, the proposed model not only guarantees the detection accuracy of the algorithm, but also greatly reduces the model volume, which provides support for the realization of visible light insulator target detection based on edge intelligence. [ABSTRACT FROM AUTHOR]

We consider the market design problem of matching students to schools in the presence of crowding effects. These effects are salient in parents' decision-making and the empirical literature; however, they cause difficulties in the design of satisfactory mechanisms and, as such, are not currently considered. We propose a new framework and an equilibrium notion that accommodates crowding, no-envy, and respect for priorities. The equilibrium has a student-optimal element that induces an incentive-compatible mechanism and is implementable via a novel algorithm. Moreover, analogs of fundamental structural results of the matching literature (the rural hospitals theorem, welfare lattice, etc.) survive. (JEL D47, H75, I21, I28) [ABSTRACT FROM AUTHOR]

Background: Head and neck squamous cell carcinoma (HNSCC) is the sixth most common cancer worldwide. A novel form of copper-dependent and reactive oxygen species (ROS)-dependent cell death, cuproptosis, has been described in many cancers. The roles and potential mechanisms of cuproptosis-related genes (CRGs) are still unclear in HNSCC. Method: We downloaded TCGA datasets of HNSCC genomic mutations and clinic data from The Cancer Genome Atlas. Based on the Cuproptosis-related differentially expressed genes in HNSCC, we constructed a prognostic signature. Results: Eight CRGs have been identified as associated with the prognosis of HNSCC. According to Kaplan–Meier analyses, HNSCC with a high Risk Score had a poor prognosis. Furthermore, the AUC of the Risk Score for the 1-, 3-, and 5- year overall survival was respectively, 0.70, 0.71, and 0.68. TCGA data revealed that T cell functions, such as HLA, cytolytic activity, inflammation regulation, co-inhibition, and co-stimulation, differed significantly between members of the low and high groups. The immune checkpoint genes PD-L1, PD-L1, and CTLA-4 were also expressed differently in the two risk groups. Conclusions: A CRG signature was defined that is associated with the prognosis of patients with HNSCC. [ABSTRACT FROM AUTHOR]

Research on batteries' State of Charge (SOC) estimation for equivalent circuit models based on the Kalman Filter (KF) framework and machine learning algorithms remains relatively limited. Most studies are focused on a few machine learning algorithms and do not present comprehensive analysis and comparison. Furthermore, most of them focus on obtaining the state space parameters of the Kalman filter frame algorithm models using machine learning algorithms and then substituting the state space parameters into the Kalman filter frame algorithm to estimate the SOC. Such algorithms are highly coupled, and present high complexity and low practicability. This study aims to integrate machine learning with the Kalman filter frame algorithm, and to estimate the final SOC by using different combinations of the input, output, and intermediate variable values of five Kalman filter frame algorithms as the input of the machine learning algorithms of six main streams. These are: linear regression, support vector Regression, XGBoost, AdaBoost, random forest, and LSTM; the algorithm coupling is lower for two-way parameter adjustment and is not applied between the machine learning and Kalman filtering framework algorithms. The results demonstrate that the integrated learning algorithm significantly improves the estimation accuracy when compared to the pure Kalman filter framework or the machine learning algorithms. Among the various integrated algorithms, the random forest and Kalman filter framework presents the highest estimation accuracy along with good real-time performance. Therefore, it can be implemented in various engineering applications. [ABSTRACT FROM AUTHOR]

In the process of two‐phase countercurrent flow in the extraction column, the study of the hydrodynamic characteristics of dispersed phase droplets is crucial for the design and scaling up of the extraction column. In a pulsed extraction column with a liquid–liquid system, the four‐sensor optical fiber probe was used to assess the hydrodynamic characteristics of the dispersed phase droplets, including size, mean velocity, and holdup. 1 M HNO3 and 30% TBP/kerosene were used as continuous phase and dispersed phase, respectively. The effects of two‐phase superficial velocities and pulsation intensity were investigated on droplet size, mean velocity, and holdup. The findings show that aside from the local holdup, the pulsation intensity has the greatest impact on the hydrodynamic characteristics. A new and concise algorithm was used to deal with the light signals to determine the droplet size and mean velocity. This algorithm created two Cartesian coordinate systems, each with a separate z‐axis using the leading probe as the origin of the coordinate instead of using the vector approach. Compared to the previous algorithm, both the needed number of the time interval between the voltage peaks and that of the direction angles decrease. The relative error of the droplet size obtained by this improved algorithm is within 7%, and droplet velocity is almost the same, compared to the previous algorithm, which indicates that the new algorithm of the four‐sensor optical fiber signal processing is reliable to the measurement of the hydrodynamic characteristics. [ABSTRACT FROM AUTHOR]

Feature selection is an important step for high-dimensional data clustering, reducing the redundancy of the raw feature set. In this paper, we focus on graph-based embedded feature selection and introduce a self-expressiveness property induced structured optimal graph feature selection (SPSOG-FS) algorithm. The proposed model incorporates both the advantages of data self-expressive property and the structured optimal graph (SOG) model. Specifically, in SPSOG-FS, a simple yet effective self-expressiveness constraint is proposed to thoroughly investigate the relations between local features. Simultaneously, we apply the SOG model to keep the original data's local structure in low-dimensional space. As a result, both feature interaction and the local structural similarity of samples are taken into consideration. In addition, we also propose an efficient method named "density peaks-based automatic clustering" (DPBAC) to estimate the number of clusters, which is a necessary prior in graph-based FS but usually unknown in real-world scenarios. According to experimental results on benchmark datasets, our proposed FS approach outperforms numerous state-of-the-art methods. Additionally, the experiments demonstrate DPBAC's capability for determining the number of clusters. [ABSTRACT FROM AUTHOR]

Background: Osteoarthritis, a common degenerative disease of articular cartilage, is characterized by degeneration of articular cartilage, changes in subchondral bone structure, and formation of osteophytes, with main clinical manifestations including increasingly serious swelling, pain, stiffness, deformity, and mobility deficits of the knee joints. With the advent of the big data era, the processing of mass data has evolved into a hot topic and gained a solid foundation from the steadily developed and improved machine learning algorithms. Aiming to provide a reference for the diagnosis and treatment of osteoarthritis, this paper using machine learning identifies the key feature genes of osteoarthritis and explores its relationship with immune infiltration, thereby revealing its pathogenesis at the molecular level. Methods: From the GEO database, GSE55235 and GSE55457 data were derived as training sets and GSE98918 data as a validation set. Differential gene expressions of the training sets were analyzed, and the LASSO regression model and support vector machine model were established by applying machine learning algorithms. Moreover, their intersection genes were regarded as feature genes, the receiver operator characteristic (ROC) curve was drawn, and the results were verified using the validation set. In addition, the expression spectrum of osteoarthritis was analyzed by immunocyte infiltration and the co-expression correlation between feature genes and immunocytes was construed. Conclusion: EPYC and KLF9 can be viewed as feature genes for osteoarthritis. The silencing of EPYC and the overexpression of KLF9 are associated with the occurrence of osteoarthritis and immunocyte infiltration. [ABSTRACT FROM AUTHOR]

Vortex electromagnetic waves (VEMWs) carrying orbital angular momentum (OAM) are promising to benefit the higher degree of freedom for inverse synthetic aperture radar (ISAR) imaging on account of their helical wavefront. For high‐speed targets, the 'stop and go' hypothesis is no longer valid. Thus, the intrapulse movement of the target will cause the broadening and shift of the one‐dimensional range profile and the blurring of the OAM eigenvalues coupling azimuth. This contribution proposes a high‐speed target imaging method for the VEMW ISAR. First, an OAM‐based ISAR imaging model for high‐speed targets is derived and the echo characteristics are analysed. Subsequently, in order to effectively estimate the ratio of the echo to correct the imaging result, an estimation method that takes into account the calculation speed and accuracy is proposed. On this basis, the target velocity distribution obtained under multiple echoes is analysed to improve the accuracy of the results. Finally, the range profile and azimuthal profile are reconstructed. Compared with the existing vortex electromagnetic imaging algorithm, the proposed method can effectively eliminate the influence of high‐speed motion on the range and azimuth profile. This work provides suggestions for the development of EM vortex ISAR imaging technology. [ABSTRACT FROM AUTHOR]

The growth of georeferenced data sources calls for advanced matching methods to improve the reliability of geospatial data processing, such as map conflation. Existing matching methods mainly focus on similarity measures at the entity scale or area scale. A measure that combines entity-scale and area-scale similarities can provide sound matching results under various circumstances. In this paper, we propose a georeferenced-graph model that integrates multiscale similarities for data matching. Specifically, a match of correspondent data objects is identified by the entity-scale measure under the constraint of the area-scale measure. Nodes in the proposed georeferenced graph model represent polygons by their centroids, whereas the links in the graph connect the nodes (i.e. centroids) according to pre-defined rules. Then, we develop an algorithm to identify many-to-many matches. We demonstrate the proposed graph model and algorithm in real-world experiments using OpenStreetMap data. The experimental results show that the proposed georeferenced-graph model can effectively integrate the context and the location-and-form distance of geospatial data matches across different datasets. [ABSTRACT FROM AUTHOR]

Battery management is the key technical link for electric vehicles. A good battery management system can realize the balanced charge and discharge of batteries, reducing the capacity degradation and the loss of health caused by battery overcharge and discharge, which all depend on the real-time and accurate estimation of the battery's state of charge (SOC). However, the battery's SOC has highly complex nonlinear time-varying characteristics related to the complex chemical and physical state and dynamic environmental conditions, which are difficult to measure directly, and this has become a difficulty in design and research. According to the characteristics of ternary lithium-ion batteries of electric vehicles, a battery SOC dual estimation algorithm based on the Variable Forgetting Factor Recursive Least Square (VFFRLS) and Multi-Innovation Unscented Kalman Filter (MIUKF) is proposed in this paper. The VFFRLS algorithm is used to estimate battery model parameters, and the MIUKF algorithm is used to estimate the battery's SOC in real time. The two algorithms are coupled to update battery model parameters and estimate the SOC. The experiment results show that the algorithm has high accuracy and stability. [ABSTRACT FROM AUTHOR]