Deep Transformation Research Articles

Deep-learning-based radiomics of intratumoral and peritumoral MRI images to predict the pathological features of adjuvant radiotherapy in early-stage cervical squamous cell carcinoma

BackgroundSurgery combined with radiotherapy substantially escalates the likelihood of encountering complications in early-stage cervical squamous cell carcinoma(ESCSCC). We aimed to investigate the feasibility of Deep-learning-based radiomics of intratumoral and peritumoral MRI images to predict the pathological features of adjuvant radiotherapy in ESCSCC and minimize the occurrence of adverse events associated with the treatment.MethodsA dataset comprising MR images was obtained from 289 patients who underwent radical hysterectomy and pelvic lymph node dissection between January 2019 and April 2022. The dataset was randomly divided into two cohorts in a 4:1 ratio.The postoperative radiotherapy options were evaluated according to the Peter/Sedlis standard. We extracted clinical features, as well as intratumoral and peritumoral radiomic features, using the least absolute shrinkage and selection operator (LASSO) regression. We constructed the Clinical Signature (Clinic_Sig), Radiomics Signature (Rad_Sig) and the Deep Transformer Learning Signature (DTL_Sig). Additionally, we fused the Rad_Sig with the DTL_Sig to create the Deep Learning Radiomic Signature (DLR_Sig). We evaluated the prediction performance of the models using the Area Under the Curve (AUC), calibration curve, and Decision Curve Analysis (DCA).ResultsThe DLR_Sig showed a high level of accuracy and predictive capability, as demonstrated by the area under the curve (AUC) of 0.98(95% CI: 0.97–0.99) for the training cohort and 0.79(95% CI: 0.67–0.90) for the test cohort. In addition, the Hosmer-Lemeshow test, which provided p-values of 0.87 for the training cohort and 0.15 for the test cohort, respectively, indicated a good fit. DeLong test showed that the predictive effectiveness of DLR_Sig was significantly better than that of the Clinic_Sig(P < 0.05 both the training and test cohorts). The calibration plot of DLR_Sig indicated excellent consistency between the actual and predicted probabilities, while the DCA curve demonstrating greater clinical utility for predicting the pathological features for adjuvant radiotherapy.ConclusionDLR_Sig based on intratumoral and peritumoral MRI images has the potential to preoperatively predict the pathological features of adjuvant radiotherapy in early-stage cervical squamous cell carcinoma (ESCSCC).

GTMicro—microservice identification approach based on deep NLP transformer model for greenfield developments

GTMicro—microservice identification approach based on deep NLP transformer model for greenfield developments

Transformers enable accurate prediction of acute and chronic chemical toxicity in aquatic organisms.

Environmental hazard assessments are reliant on toxicity data that cover multiple organism groups. Generating experimental toxicity data is, however, resource-intensive and time-consuming. Computational methods are fast and cost-efficient alternatives, but the low accuracy and narrow applicability domains have made their adaptation slow. Here, we present a AI-based model for predicting chemical toxicity. The model uses transformers to capture toxicity-specific features directly from the chemical structures and deep neural networks to predict effect concentrations. The model showed high predictive performance for all tested organism groups-algae, aquatic invertebrates and fish-and has, in comparison to commonly used QSAR methods, a larger applicability domain and a considerably lower error. When the model was trained on data with multiple effect concentrations (EC50/EC10), the performance was further improved. We conclude that deep learning and transformers have the potential to markedly advance computational prediction of chemical toxicity.

Inversion of Soil Salinity in the Irrigated Region along the Southern Bank of the Yellow River Using UAV Multispectral Remote Sensing

Inversion of Soil Salinity in the Irrigated Region along the Southern Bank of the Yellow River Using UAV Multispectral Remote Sensing

Vocational Education Metaverse: Research Status, Existing Problems, and Development Strategies

The digital strategic action has promoted the high-quality development of vocational education. As an important measure for the digital transformation of vocational education, the application and development of educational metaverse are related to the deep transformation of vocational education and the overall development of smart education. This study first reviews the current research status of the educational metaverse in vocational education from the aspects of technological development, application status, and implementation effects. Secondly, it points out the current practical problems in vocational education metaverse from the perspectives of infrastructure, technical standards system, practical training teaching model, and ethical risks. Finally, it proposes targeted improvements from the following four aspects: first, improve the infrastructure of vocational education metaverse; second, accelerate the establishment of technical standards for vocational education metaverse; explore teaching models and systems of vocational education metaverse; fourth, effectively prevent ethical risks of vocational education metaverse.

Books, videos and platforms: Exploring the BookTube interface

The publishing market has gone through deep transformations in the last few years, and the emergence of new media actors like BookTubers has increased the complexity of this system. The article proposes understanding BookTuber practices based on an interface-centered approach that goes beyond the conventional studies of the “graphic user interface”. In the framework employed in the article, the “interface” is considered to be the place where different actors relate to each other and the process between them are carried out. The article also presents a map of the BookTuber interface and its tensions. This interface-centered approach, a work-in-progress, could also be applied to other practices.

Temporal cluster-based local deep learning or signal processing-temporal convolutional transformer for daily runoff prediction?

Water scarcity poses a major obstacle to sustainable development, and precise discharge prediction plays a vital role in enabling effective water resource management. This study investigated improved prediction techniques for nonstationary time series. The study evaluated the effect of signal processing techniques and blending approaches on the performance of deep learning models for daily discharge prediction. It also compared the performance of cluster-based local modeling with hybrid signal processing-deep learning approaches. Two robust deep learning methods, Long Short-Term Memory (LSTM) and Convolutional Neural Network (CNN), along with a powerful signal processing approach called discrete wavelet transform, were utilized for prediction of daily discharge. Three blending approaches were assessed: 1) decomposing both inputs and target, 2) decomposing only the target, and 3) decomposing only the inputs and then blending them with deep learning models. Also, a new hybrid deep learning based model namely discrete wavelet transform-Temporal Convolutional Transformer (DWT-TCT) was developed. The results showed that a single-output wavelet transform-deep learning model (3rd blending approach) outperformed multi-output models, demonstrating a relative enhancement of up to 56% for the LSTM model and 51% for the CNN model. Furthermore, temporal cluster-based local modeling displayed promising performance, resulting in an improvement of up to 18% in NRMSE compared to the wavelet transform-deep learning model, while also requiring less computational cost. The successful results of the temporal cluster-based local modeling approach provide a beneficial alternative to hybrid signal processing-deep learning models. In addition the results showed that the proposed DWT-TCT model outperformed all other models with NRMSE ranges from 6.8% to 16.2% in the study areas. The results have implications for hydrology and water resources management, as they can be used to develop more precise and effective models for predicting discharge in view of nonstationarity.

Hydrothermal transformation of fly ash to tobermorite or katoite: Impact of Ca and Si concentration in the liquid phase without alkali activation

In the study of hydrothermal conversion of fly ash to calcium silicate hydrates (C–S–H), the concentration of Ca and Si in the liquid phase plays a vital role in determining the hydrothermal transformation results. This paper selected a fly ash as a research object, which can leach a high Si concentration into the solution in the hydrothermal process even in a slight alkaline solution, based on a new finding when studying the alkali dissolution kinetics of different fly ashes. To explore the effects of Ca and Si concentrations in the liquid phase on the hydrothermal transformation of fly ash to tobermorite or katoite, Na2SiO3 (Na2O·xSiO2·yH2O) and CaSO4 were used to change the concentration of Si and Ca in the liquid in the hydrothermal system. Characterization methods such as XRD, SEM, infrared spectroscopy, and solid-state nuclear magnetic resonance indicated that a little more concentration of Si can promote the transformation from fly ash to tobermorite and persuade the conversion of the Si–O tetrahedron in C–S–H from Q2 to Q3 and Q4. However, too high a liquid Si concentration made the hydrothermal product change to colloidal calcium silicate. On the other hand, a high Ca concentration in the liquid was found to lead to the conversion of the hydrothermal product from tobermorite to katoite. According to the above results, the deep transformation from fly ash to tobermorite under the conditions of a low temperature of 160 °C and no extra alkali addition was achieved by the induction of a small quantity of Na2SiO3.

Deep attention transformer nets for accurate analysis of oil spilled images to minimize pollution in the marine environment

Oil spills in maritime areas pose a serious environmental risk, wreaking havoc on marine ecosystems, coastal habitats, and local residents. An accurate and timely evaluation of oil spill occurrences and extent is critical for effective pollution control and mitigation. In this study, we present a novel and cutting-edge approach for analyzing oil-spilled images using Deep Attention Transformer Nets (DATN) with Collective Intelligence (CI), with the goal of reducing pollution in the marine environment. This method takes advantage of deep learning capability, notably the incorporation of transformer-based attention processes, to improve the identification and measurement of oil spills in satellite and aerial images. The DATN model is intended to learn complicated features from images automatically, capturing complex patterns associated with oil spills and their surrounding context. The model chooses focus on key regions and add spatial links by using attention mechanisms, allowing for a more comprehensive understanding of the environmental influence. We thoroughly test DATN performance using a variety of datasets encompassing various oil spill scenarios and environmental circumstances. The results show that DATN surpasses standard approaches and other deep learning models in recognizing oil spill regions, with excellent accuracy, precision, and recall rates. Furthermore, the model has strong generalization capabilities across a wide range of image sources and situations.

An advanced approach to operational quality evaluation for industry-finance integration enterprises based on integrated interval-valued intuitionistic fuzzy multi-attribute decision making

With the acceleration of economic globalization, enterprises are facing fierce competition and huge challenges, requiring deep financial management transformation. In this context, the integration of industry and finance has gradually demonstrated extremely important value. The integration of industry and finance can not only effectively improve the efficiency of financial management, prevent business risks, and improve operational efficiency, but also enhance the comprehensive ability of enterprise financial management, providing a more flexible, transparent, and efficient financial management system for enterprises. The operational quality evaluation of industry-finance integration enterprises under lean management accounting is a multiple-attribute decision-making (MADM). In this paper, some calculating laws on IVIFSs, Hamacher sum, Hamacher product are introduced, and the interval-valued intuitionistic fuzzy Hamacher interactive power averaging (IVIFHIPA) technique is proposed based on the interval-valued intuitionistic fuzzy (IVIF) Hamacher interactive weighted averaging (IVIFHIWA) technique and power average (PA) technique. Meanwhile, some ideal properties of IVIFHIPA technique are studied. Then, the IVIFHIPA technique is employed to cope with the MADM under IVIFSs. Finally, an example for operational quality evaluation of industry-finance integration enterprises under lean management accounting is employed to test the IVIFHIPA technique. Thus, the main research aim of this paper is concluded as follows: (1) the IVIFHIPA technique is constructed based on IVIFHIWA technique and classical power average (PA) technique; (2) the IVIFHIPA technique is put forward to cope with the MADM under IVIFSs; (3) an empirical example for operational quality evaluation of industry-finance integration enterprises under lean management accounting has been put forward to show the IVIFHIPA technique.

Is Taylor Swift leading a new Pop revolution? A cross-generation analysis of Pop/Rock cover songs

Background (Pop/Rock is experiencing deep transformations caused by the emergence of Web 2.0 and the online social networks, streaming services and the proliferation of TV contests. Cover songs afford the opportunity to explore how these changes would result in a new revalorization of Pop and Dance music. The aim of this paper is to find evidence of a changing pattern in cover songs by 21st-century artists. To this end, over 76,000 covers performed by artists who grew up in the 2000s were quantitatively analyzed. Methods SecondHandSongs.com was crawled to extract the cover relationships, and Allmusic.com to obtain the genre and starting decade of each performer. Results The results show that the current music panorama is dominated by Pop/Rock music and, for the first time since the 1960s, artists from the 2010s generation prefer to cover more songs by contemporary artists than by classic figures from the 1960s. Pop and Dance are the emerging sub-genres with the largest proportion of covered musicians, while Taylor Swift, Justin Bieber and Ed Sheeran are responsible for this changing trend. Conclusions These results provide an interesting opportunity to introduce quantitative studies in cultural studies about music, cinema and arts.

Neural network enabled nanoplasmonic hydrogen sensors with 100 ppm limit of detection in humid air

Environmental humidity variations are ubiquitous and high humidity characterizes fuel cell and electrolyzer operation conditions. Since hydrogen-air mixtures are highly flammable, humidity tolerant H2 sensors are important from safety and process monitoring perspectives. Here, we report an optical nanoplasmonic hydrogen sensor operated at elevated temperature that combined with Deep Dense Neural Network or Transformer data treatment involving the entire spectral response of the sensor enables a 100 ppm H2 limit of detection in synthetic air at 80% relative humidity. This significantly exceeds the <1000 ppm US Department of Energy performance target. Furthermore, the sensors pass the ISO 26142:2010 stability requirement in 80% relative humidity in air down to 0.06% H2 and show no signs of performance loss after 140 h continuous operation. Our results thus demonstrate the potential of plasmonic hydrogen sensors for use in high humidity and how neural-network-based data treatment can significantly boost their performance.

Distributed Split Learning for Map-Based Signal Strength Prediction Empowered by Deep Vision Transformer

Distributed Split Learning for Map-Based Signal Strength Prediction Empowered by Deep Vision Transformer

APT Attack Detection of a New Power System based on DPI-transformer

Introduction: In recent years, the frequent occurrence of network security attacks in the power field has brought huge risks to the production, transmission, and supply of power systems, and Advanced Persistent Threat (APT) is a covert advanced network security attack, which has become one of the network security risks that cannot be ignored in the construction of new power systems. Objective: This study aims to resist the increasing risk of APT attacks in the construction of new power systems, this paper proposes an attack detection model based on Deep Packet Inspection (DPI) and Transformer. Methods: Firstly, we extracted 606 traffic characteristics from the original traffic data through the extended CIC Flowmeter and used them all to train the Transformer network. Then, we used the DPI-Transformer model and traffic labels to perform feature analysis on the traffic data and finally obtained the APT-Score. If the APT-Score is greater than the threshold, the alarm module is triggered. Results: By analyzing the headers and payloads of the network traffic in the APT-2020 dataset, the experimental results show that the detection accuracy of APT attacks by the DPI-Transformer detection model is significantly higher than that of the current mainstream APT attack detection algorithms. Conclusion: Combined with the characteristics of the new power system and APT attacks, this paper proposes an attack detection model DPI-Transformer, which proves that the model has greatly improved the detection accuracy.

Residual Shrinkage ViT with Discriminative Rebalancing Strategy for Small and Imbalanced Fault Diagnosis

In response to the challenge of small and imbalanced Datasets, where the total Sample size is limited and healthy Samples significantly outweigh faulty ones, we propose a diagnostic framework designed to tackle Class imbalance, denoted as the Dual-Stream Adaptive Deep Residual Shrinkage Vision Transformer with Interclass–Intraclass Rebalancing Loss (DSADRSViT-IIRL). Firstly, to address the issue of limited Sample quantity, we incorporated the Dual-Stream Adaptive Deep Residual Shrinkage Block (DSA-DRSB) into the Vision Transformer (ViT) architecture, creating a DSA-DRSB that adaptively removes redundant signal information based on the input data characteristics. This enhancement enables the model to focus on the Global receptive field while capturing crucial local fault discrimination features from the extremely limited Samples. Furthermore, to tackle the problem of a significant Class imbalance in long-tailed Datasets, we designed an Interclass–Intraclass Rebalancing Loss (IIRL), which decouples the contributions of the Intraclass and Interclass Samples during training, thus promoting the stable convergence of the model. Finally, we conducted experiments on the Laboratory and CWRU bearing Datasets, validating the superiority of the DSADRSViT-IIRL algorithm in handling Class imbalance within mixed-load Datasets.

Extracting Drug-Protein Relation from Literature Using Ensembles of Biomedical Transformers.

Automatic extraction of relations between drugs/chemicals and proteins from ever-growing biomedical literature is required to build up-to-date knowledge bases in biomedicine. To promote the development of automated methods, BioCreative-VII organized a shared task - the DrugProt track, to recognize drug-protein entity relations from PubMed abstracts. We participated in the shared task and leveraged deep learning-based transformer models pre-trained on biomedical data to build ensemble approaches to automatically extract drug-protein relation from biomedical literature. On the main corpora of 10,750 abstracts, our best system obtained an F1-score of 77.60% (ranked 4th among 30 participating teams), and on the large-scale corpus of 2.4M documents, our system achieved micro-averaged F1-score of 77.32% (ranked 2nd among 9 system submissions). This demonstrates the effectiveness of domain-specific transformer models and ensemble approaches for automatic relation extraction from biomedical literature.

Controlling Media Player with Hands: A Transformer Approach and a Quality of Experience Assessment

In this article, we propose a Hand Gesture Recognition (HGR) system based on a novel deep transformer (DT) neural network for media player control. The extracted hand skeleton features are processed by separate transformers for each finger in isolation to better identify the finger characteristics to drive the following classification. The achieved HGR accuracy (0.853) outperforms state-of-the-art HGR approaches when tested on the popular NVIDIA dataset. Moreover, we conducted a subjective assessment involving 30 people to evaluate the Quality of Experience (QoE) provided by the proposed DT-HGR for controlling a media player application compared with two traditional input devices, i.e., mouse and keyboard. The assessment participants were asked to evaluate objective (accuracy) and subjective (physical fatigue, usability, pragmatic quality, and hedonic quality) measurements. We found that (i) the accuracy of DT-HGR is very high (91.67%), only slightly lower than that of traditional alternative interaction modalities; and that (ii) the perceived quality for DT-HGR in terms of satisfaction, comfort, and interactivity is very high, with an average Mean Opinion Score (MOS) value as high as 4.4, whereas the alternative approaches did not reach 3.8, which encourages a more pervasive adoption of the natural gesture interaction.

Novelty fused image and text models based on deep neural network and transformer for multimodal sentiment analysis

Novelty fused image and text models based on deep neural network and transformer for multimodal sentiment analysis

Developing Encounters between Chinese and Western Art: The Contribution of Two Jesuits in China in the Seventeenth and Eighteenth Centuries

This article explores the introduction and influence of Western art in China during the Ming and Qing dynasties, focusing on the role of the Jesuits—especially Matteo Ricci (1552–1610), one of the founders of Catholic missionary work in China—and the most famous and influential Western figure in China’s art history, Giuseppe Castiglione (1688–1766). The Jesuits recognized art’s potential in their missionary efforts. The article examines the varied responses of Chinese literati to Western art theories in the 17th and 18th centuries, and how Jesuit-introduced artistic concepts were assimilated within different Chinese contexts. It also investigates how Western art merged into Chinese culture, noting the linguistic integration of concepts like chiaroscuro. In Ricci’s time, Western art intrigued the Chinese, but deep acceptance and conceptual transformation were limited. The assimilation of Western techniques was not as widespread as expected. In the 18th century, due to restrictions on Christianity, the religious role of painting diminished. Castiglione’s hybrid style, though influential in the Qing court, faced challenges in gaining cultural acceptance from the Han literati, leading to more criticism than in Ricci’s era. Despite these challenges and varying receptions, the contributions of Ricci and Castiglione to the Sino–Western art exchange are significant, reflecting the complex interplay of art, religion, and cultural dynamics across these periods.

Męczeństwo Sługi Bożego Jánosa Esterházyego i tego dzisiejsza interpretacja

The text discusses the fate of János Esterházy, an aristocrat and activist of the Hungarian minority in Czechoslovakia. Arrested in 1945 and deported to the Soviet Union, he returned to Czechoslovakia in 1949. Pardoned from the death sentence, he spent the rest of his life in prison, where he died of exhaustion in 1957. The essential part of his life (?) is his deep inner spiritual transformation, which took place in the most difficult period of his life. It led to the search for the meaning of suffering in the cross of Christ. Its fruit was a steadfast attitude of forgiveness and reconciliation in the spirit of mercy, confirmed by numerous accounts.