Selective Embedding Research Articles

Prospective selective embedding of radical prostatectomy specimens is not inferior to full embedding regarding established and new prognostic parameters.

The histopathological examination of radical prostatectomy specimens is essential for assessing critical tumor characteristics, including stage, grade, and margins, all of which impact patient prognosis. However, the extent of embedding the prostate has long been a subject of debate, with some advocating partial/selective embedding and others favoring complete embedding. This study establishes a standardized and time-efficient protocol for processing radical prostatectomy specimens with limited embedding while maintaining diagnostic accuracy. Two hundred twenty-six prostatectomy specimens were analyzed, and the results of a highly standardized selective embedding protocol, systematically embedding the apex, the base, the transition to the seminal vesicles, and selected horizontal sections, were compared with full embedding as the gold standard. Non-inferiority testing was conducted by one-sided binomial tests and Pearson-Clopper confidence intervals. Selective embedding provided consistent and accurate diagnostic information with up to 90-98% concordance in pT, margins, ISUP-grade groups, and presence of IDC-P and cribriform tumor growth. In summary, this study establishes an economical standardized protocol for selective embedding of radical prostatectomy specimens with only minimal loss of information.

Glycoprobe for selective embedding in membrane rafts

Glycoprobe for selective embedding in membrane rafts

ASENN: attention-based selective embedding neural networks for road distress prediction

This study proposes an innovative neural network framework, ASENN (Attention-based Selective Embedding Neural Network), for the prediction of pavement deterioration. Considering the complexity and uncertainty associated with the pavement deterioration process, two fundamental frameworks, SEL (Selective Embedding Layer) and MDAL (Multi-Dropout Attention Layer), are combined to enhance feature abstraction and prediction accuracy. This approach is significant while analyzing the pavement deterioration process due to the high variability of the contributing deterioration factors. These factors, represented as tabular data, undergo filtering, embedding, and fusion stages in the SEL, to extract crucial features for an effective representation of pavement deterioration. Further, multiple attention-weighted combinations of raw data are obtained through the MDAL. Several SELs and MDALs were combined as basic cells and layered to form an ASENN. The experimental results demonstrate that the proposed model outperforms existing tabular models on four road distress parameter datasets corresponding to cracking, deflection, international roughness index, and rutting. The optimal number of cells was determined using different ablation settings. The results also show that the feature learning capabilities of the ASENN model improved as the number of cells increased; however, owing to the limited combination space of feature fields, extreme depths were not preferred. Furthermore, the ablation investigation demonstrated that MDAL can improve performance, particularly on the cracking dataset. Notably, compared with mainstream transformer models, ASENN requires significantly less storage and achieves faster execution speed.

Liquid sculpture and curing of bio-inspired polyelectrolyte aqueous two-phase systems

Aqueous two-phase systems (ATPS) provide imperative interfaces and compartments in biology, but the sculpture and conversion of liquid structures to functional solids is challenging. Here, inspired by phase evolution of mussel foot proteins ATPS, we tackle this problem by designing poly(ionic liquids) capable of responsive condensation and phase-dependent curing. When mixed with poly(dimethyl diallyl ammonium chloride), the poly(ionic liquids) formed liquid condensates and ATPS, which were tuned into bicontinuous liquid phases under stirring. Selective, rapid curing of the poly(ionic liquids)-rich phase was facilitated under basic conditions (pH 11), leading to the liquid-to-gel conversion and structure sculpture, i.e., the evolution from ATPS to macroporous sponges featuring bead-and-string networks. This mechanism enabled the selective embedment of carbon nanotubes in the poly(ionic liquids)-rich phase, which showed exceptional stability in harsh conditions (10 wt% NaCl, 80 oC, 3 days) and high (2.5 kg/m2h) solar thermal desalination of concentrated salty water under 1-sun irradiation.

Solution-processable electrode-material embedding in dynamically inscribed nanopatterns (SPEEDIN) for continuous fabrication of durable flexible devices

A facile and scalable lithography-free fabrication technique, named solution-processable electrode-material embedding in dynamically inscribed nanopatterns (SPEEDIN), is developed to produce highly durable electronics. SPEEDIN uniquely utilizes a single continuous flow-line manufacturing process comprised of dynamic nanoinscribing and metal nanoparticle solution coating with selective embedding. Nano- and/or micro-trenches are inscribed into arbitrary polymers, and then an Ag nanoparticle solution is dispersed, soft-baked, doctor-bladed, and hard-baked to embed Ag micro- and nanowire structures into the trenches. Compared to lithographically embossed metal structures, the embedded SPEEDIN architectures can achieve higher durability with comparable optical and electrical properties and are robust and power-efficient even under extreme stresses such as scratching and bending. As one tangible application of SPEEDIN, we demonstrate a flexible metal electrode that can operate at 5 V at temperatures up to 300 °C even under the influence of harsh external stimuli. SPEEDIN can be applied to the scalable fabrication of diverse flexible devices that are reliable for heavy-duty operation in harsh environments involving high temperatures, mechanical deformations, and chemical hazards.

An Adaptive Unsupervised Feature Selection Algorithm Based on MDS for Tumor Gene Data Classification.

Identifying the key genes related to tumors from gene expression data with a large number of features is important for the accurate classification of tumors and to make special treatment decisions. In recent years, unsupervised feature selection algorithms have attracted considerable attention in the field of gene selection as they can find the most discriminating subsets of genes, namely the potential information in biological data. Recent research also shows that maintaining the important structure of data is necessary for gene selection. However, most current feature selection methods merely capture the local structure of the original data while ignoring the importance of the global structure of the original data. We believe that the global structure and local structure of the original data are equally important, and so the selected genes should maintain the essential structure of the original data as far as possible. In this paper, we propose a new, adaptive, unsupervised feature selection scheme which not only reconstructs high-dimensional data into a low-dimensional space with the constraint of feature distance invariance but also employs -norm to enable a matrix with the ability to perform gene selection embedding into the local manifold structure-learning framework. Moreover, an effective algorithm is developed to solve the optimization problem based on the proposed scheme. Comparative experiments with some classical schemes on real tumor datasets demonstrate the effectiveness of the proposed method.

Reversible selective embedding for DICOM image security and integrity using visual cryptography

Information security has been one of the biggest challenges in the era of telemedicine applications. To protect the patient's private records in medical images, several traditional methods are used. One of them is hiding the sensitive part of the image by embedding a different message into it. The data embedding can be done using methods like reversible data hiding (RDH) or reversible watermarking in the region of interest (ROI) of the original medical images. For ensuring the authenticity and integrity of the received medical image, an innovative data embedding technique is proposed in this paper. The proposed technique chooses selective blocks from the ROI based on the smoothness function and the region of non-interest (RONI) is embedded into ROI using some reversible operations. The proposed technique is an application of visual cryptography where the embedded region cannot be visually detected by an attacker. The experimental results prove the reversibility and the efficiency of the proposed approach.

Heterobimetallic Au(i)/Y(iii) single chain nanoparticles as recyclable homogenous catalysts

Au(i)/Y(iii) single chain nanoparticles (SCNPs) are potent homogenous, recyclable catalysts for the hydroamination. The SCNPs consist of terpolymer chains with orthogonal ligand units, enabling the selective embedding of different metals.

Gemma: Reinforcement Learning-Based Graph Embedding and Mapping for Virtual Network Applications

Virtual network mapping (VNM) is a challenge in the field of network virtualization. As VNM variants have been formalized depending on substrate network structures, virtual network specifications, mapping optimization objectives, and other factors, a number of VNM heuristic methods have been introduced. On the other hand, reinforcement learning (RL) algorithms have been incorporated into deep learning frameworks and recognized as a promising solution for solving complex resource allocation problems. In this paper, we present an RL-based graph embedding and mapping framework, Gemma, for tackling various VNM problems in a unified end-to-end manner. In the framework, we employ an encoder-decoder deep learning architecture and propose several optimization schemes such as two-stage mapping and model-based selective embedding. Aiming to deal with large-scale VNM problems in both online and offline scheduling systems, the proposed schemes explore the trade-off between inference accuracy and mapping function runtimes, enhancing scalability and timeliness. Gemma shows robust performance under various problem conditions, outperforming other heuristic and learning-based methods.

A Comparative Study of Sentiment Analysis on Mask-Wearing Practices during the COVID-19 Pandemic

COVID-19 has become one of the most highly orated subject matter in these days. Countries have taken many viable actions to prevent the spread of the virus directed by international recommendations, which led to many disputes concerning wearing a face mask as a preventive measure against the virus. This study aims to assess and compare the overall accuracy, macro precision, macro F-measure and macro recall of the different decision models towards the COVID-19 mask-wearing practices via sentiment analysis. Tweets are labeled and text pre-processing techniques are applied as stemming, normalization, tokenization, and stop-word removal. Subsequently, the tweets are transformed into master feature vectors by applying various feature extraction, feature representation, feature selection and word embedding techniques with five supervised machine learning decision models to predict mask wearing practices reinforced from Twitter tweets. Moreover, the highest macro F-measure and macro precision are found with feature extraction as hybrid-grams, feature representation as TF-IDF, feature selection as Chi-Squared Test, and highest macro recall with feature extraction as BOW, feature representation as TF-IDF, feature selection as ANOVA F-value. Hence, this study concludes that the Naive Bayes (NB) algorithm outperforms other decision models with master feature vectors applied. In addition, it also outperforms word embedding techniques.

Hybrid DNA Chaos Based Steganography on ECG Signal for Secured Data Communication

One of the principal issues in data sharing is security. Data being transferred on the internet is prone to severe attacks. Securing data is kind of digital privacy measures that are implemented to prevent the data to be hacked by unauthorized person while sharing the information over the communication network. Data may be of any form like image, signal or text. Encrypting the data before being transmitted is one step taken to promote security which is the well-known process of cryptography. Steganography deals with hiding the message in any unknown format. Combination of these methods turns to be more secure. A hybrid DNA hyper chaotic based algorithm with lifting wavelet technology to secure ECG signal is proposed here. The wavelet technique that decomposes the image into four bands namely LL, LH, HH, HL. To remove the redundancies the coefficients are utilised in the encoder process. The data hiding is implemented by selective embedding process is introduced in this method by determine host signal samples. The encrypted data then is hidden inside image using DNA method logically. When one knows and tries to get the hidden data they require the secret key to do so and any mistake in that leads to distortion of the signal which results in loss of data. This method is found to be secure and the simulation is done using MATLAB. A GUI is developed to choose the required ECG signal and other operations to embed the data and find the original signal after retrieval.

Exploiting De-Noising Convolutional Neural Networks DnCNNs for an Efficient Watermarking Scheme: A Case for Information Retrieval

ABSTRACT Digital watermarking has various applications in which it plays an important role; these include document security and multimedia vendor copyrights. This paper presents an efficient watermarking technique called ‘modified selective embedding in low frequency’ (M-SELF) implementing a Fast Fourier Transform (FFT) for frequency domain analysis where FFT shift brings down lower-frequency components to the center of the cover image. Embedding of the watermark is done by optimal radial implementation after coordinates computation via phase and magnitude separation. Furthermore, the watermark is protected by the convolutional code (1/2) and Advanced Encryption Standard (AES-128) algorithm providing a double layer of security. Moreover, the exploitation of de-noising convolutional neural networks (DnCNNs) is another contribution of the proposed work. The network is utilized as a de-noiser in an integrated manner for the proposed watermarking technique on blind level of additive white Gaussian noise (AWGN) to check the robustness of the technique. Performance evaluation of the proposed integrated watermarking technique is done in terms of percentage of bits being retrieved as a success rate at each level of noise and perceived visual quality by using full-reference image quality assessment (IQA) metrics such as peak signal-to-noise ratio (PSNR) and structural similarity index (SSIM).

A high capacity reversible data hiding through multi-directional gradient prediction, non-linear regression analysis and embedding selection

The technique of reversible data hiding enables an original image to be restored from a stego-image with no loss of host information, and it is known as a reversible data hiding algorithm (RDH). Our goal is to design a method to predict pixels effectively, because the more accurate the prediction is, the more concentrated the histogram is, and it minimizes shifting to avoid distortion. In this paper, we propose a new multi-directional gradient prediction method to generate more accurate prediction results. In embedding stage, according to the embedding capacity of information, we generate the best decision based on non-linear regression analysis, which can differentiate between embedding region and non-embedding region to reduce needless shifting. Finally, we utilize the automatic embedding range decision. With sorting by the amount of regional variance, the easier predicted region is prior for embedding, and the quality of the image is improved after embedding. To evaluate the proposed reversible hiding scheme, we compared other methods on different pictures. Results show that the proposed scheme can embed much more data with less distortion.

Selective Embedding with Gated Fusion for 6D Object Pose Estimation

Deep learning method for 6D object pose estimation based on RGB image and depth (RGB-D) has been successfully applied to robot grasping. The fusion of RGB and depth is one of the most important difficulties. Previous works on the fusion of these two features are mostly concatenated together without considering the different contributions of the two types of features to pose estimation. We propose a selective embedding with gated fusion structure called SEGate, which can adjust the weights of RGB and depth features adaptively. Furthermore, we aggregate the local features of point clouds according to the distance between them. More specifically, the close point clouds contribute a lot to local features, while the distant point clouds contribute a little. Experiments show that our approach achieves the state-of-art performance in both LineMOD and YCB-Video datasets. Meanwhile, our approach is more robust to the pose estimation of occluded objects.

A new graph-preserving unsupervised feature selection embedding LLE with low-rank constraint and feature-level representation

Unsupervised feature selection is a powerful tool to process high-dimensional data, in which a subset of features is selected out for effective data representation. In this paper, we proposes a novel robust unsupervised features selection method based on graph-preserving feature selection embedding LLE. Specifically, we integrate the graph matrix learning and the low-dimensional space learning together to identify the correlation among both features and samples from the intrinsic low-dimensional space of original data. Also, the global and local correlation of features have been taken into consideration through the low-rank constraint and the feature-level representation property to find lower-dimensional representation which preserves not only the global and local correlation of features but also the global and local structure of training samples. Furthermore, we propose a new optimization algorithm to the resulting objective function, which iteratively updates the graph matrix and the intrinsic space in order to collaboratively improve each of them. Experimental analysis on 18 benchmark datasets verified that our proposed method outperformed the state-of-the-art feature selection methods in terms of classification and clustering performance.

LJELSR: A Strengthened Version of JELSR for Feature Selection and Clustering.

Feature selection and sample clustering play an important role in bioinformatics. Traditional feature selection methods separate sparse regression and embedding learning. Later, to effectively identify the significant features of the genomic data, Joint Embedding Learning and Sparse Regression (JELSR) is proposed. However, since there are many redundancy and noise values in genomic data, the sparseness of this method is far from enough. In this paper, we propose a strengthened version of JELSR by adding the L1-norm constraint on the regularization term based on a previous model, and call it LJELSR, to further improve the sparseness of the method. Then, we provide a new iterative algorithm to obtain the convergence solution. The experimental results show that our method achieves a state-of-the-art level both in identifying differentially expressed genes and sample clustering on different genomic data compared to previous methods. Additionally, the selected differentially expressed genes may be of great value in medical research.

Shape Adaptable Medical Multimedia Processing

Introduction:Electronic medical healthcare systems are becoming the backbone of health organizations over the world. The huge amounts of medical multimedia produced by these systems especially images and videos are transmitted by the computer networks that connect these systems. The variability in the shape and texture of transferred medical multimedia data needs adaptable procedures to process these data efficiently. In other words, these procedures must adjust automatically based on the shape of region of interests in the medical multimedia images to cope with fast changes in the healthcare environments. In this paper, we have proposed shape adaptable watermarking approaches for medical multimedia processing systems. The medical images generated by X-rays, MRI and CT modalities are used in our experiments to test proposed approaches. In addition, these approaches were tested under different kinds of signal processing and geometric attacks. The comparative comparison of our proposed approaches with state-of-art approaches proved the superiority and capability of our approaches to adjust the number of selected subands of medical cover image to embed and extract the hospital watermark logos.Background and Objective:The development of an adjustable approach to process medical multimedia signals for healthcare system. The aim of this research is to select adaptably the number of subands of cover image to hide the information of hospital logo watermark inside them such that embedded watermarks can resist different kinds of attacks.Method:The proposed adjustable approach consists of suband selection method, criterion, embedding and extraction procedures, DWT transform, attacks, evaluation metrics,etc.Results & Conclusion:It provides robust and adjustable method to embed and extract watermark logo at different resolution levels of cover medical image and uses with images of different sizes and modalities.

Selective embedment of silver nanocrystals into spatially segregated domains in thin polymer films for controlled fabrication of functional nanocomposites.

Fabrication of polymer-nanoparticle nanocomposites typically relies on mixing nanoparticle and polymer solutions, which renders little control over nanoparticle incorporation, and homogeneity of the resulting composite material. This work focuses on the thermally induced embedment of monocrystalline silver nanocubes (AgNCs) into polymer surfaces. The AgNCs are initially deposited through a Langmuir approach onto films of immiscible blended polymer films, which allows fine control over nanoparticle density and aggregation state. This nanoparticle/polymer composite is then heated above the glass transition temperature (Tg) of a polymer, which initiates the irreversible embedding of the AgNCs. The immiscible ternary polymer films featured discrete domains (with different Tgs), which were altered by changing the amount of polystyrene, poly(2-vinylpyridine) and poly(methyl methacrylate) within the polymer solution. The Tg dependence of the embedding process allowed the selective embedment of AgNCs into discrete polymer domains. The process was monitored in real time by using spatially separated hybrid plasmon modes, through peak shifts observed in a UV-vis spectrum. Enhanced surface confinement was observed for certain tripolymer films when compared to polystyrene–AgNC nanocomposites, due to changes in the surface energy within the blend. This work brings interesting insight on nanoparticle-blended polymer interactions and provides a fairly universal approach for the fabrication of these polymer–metal nanoparticle nanocomposites, which is of particular interest in fields that require fine control over nanoparticle incorporation within segregated polymer domains.

Spatial-time-state fusion algorithm for defect detection through eddy current pulsed thermography

Eddy Current Pulsed Thermography (ECPT) has received extensive attention due to its high sensitive of detectability on surface and subsurface cracks. However, it remains as a difficult challenge in unsupervised detection as to identify defects without knowing any prior knowledge. This paper presents a spatial-time-state features fusion algorithm to obtain fully profile of the defects by directional scanning. The proposed method is intended to conduct features extraction by using independent component analysis (ICA) and automatic features selection embedding genetic algorithm. Finally, the optimal feature of each step is fused to obtain defects reconstruction by applying common orthogonal basis extraction (COBE) method. Experiments have been conducted to validate the study and verify the efficacy of the proposed method on blind defect detection.

Efficient lossless watermarking algorithm using gradient sorting and selective embedding

Prediction-error expansion based lossless watermarking techniques for images have received extensive attention due to versatile performance. In particular, the capacity-distortion control is one of the most important studies. This paper investigates that data-hiding distortion is mainly affected by the occurrence order of the prediction-error distribution, and capacity is affected by the occurrence number of the prediction errors. To make better control of the prediction-error distortion, we propose a novel scheme with three components: the gradient adjacent prediction (GAP), gradient sorting, and selective embedding. The capacity of the proposed scheme is high because GAP generates highly concentrated prediction-error histogram. Furthermore, the gradient sorting and selective embedding could avoid and reduce the unnecessary distortion. Experimental results show that the proposed scheme outperforms the competitors in terms of capacity enhancement and distortion reduction.