Transformation Framework Research Articles

Lightweight multimodal Cycle-Attention Transformer towards cancer diagnosis

Cancer diagnosis, prognosis and therapeutic response predictions are based on data extracted from various modalities, such as histology slides, cell graphs and molecular profiles from genomic data. In recent years, with the increasing availability of multimodal biomedical data, an increasing number of multimodal fusion strategies have been proposed to overcome the challenge of data heterogeneity and make full use of intersections and complementarities of different data sources. However, most of them have the problem of parameter redundancy and inefficiency in the process of fusion. Therefore, in this work, we present a Cycle-Attention Transformer (Cy-Atten) framework that learns a more robust joint representation for multimodal cancer survival analysis. The Cy-Atten adopts a cycling mechanism with sparse attention and yields a theoretical cost of O(N) for adding a modality, which is proven to effectively reduce the parameter redundancy. Empirically in the experiments on two cancer datasets from the TCGA database, the parameters in our fusion part are only of the order of 103 considering three modalities. In addition to being lightweight, the Cy-Atten outperforms several state-of-the-art multimodal fusion methods in survival analysis and grade classification of two cancers from the TCGA database. Our work can be reproduced at https://github.com/dlmscl/Cycle-Attention-Transformer.

A novel method for aero-engine time-series forecasting based on multi-resolution transformer

Time-series forecasting is widely studied in the data-driven component-level modeling, fault diagnosis, and performance prediction of aero-engines. The operational data of aero-engines have the characteristics of complexity and nonlinearity. This paper proposes a multi-resolution transformer (MRT) model for the non-steady state process of aero-engines. This transformer-based model can learn temporal patterns of different scales by performing multi-resolution down-sampling and patch-based tokeniczation on the input time-series. On this basis, a two-stage multi-resolution transformer (TSMRT) framework is proposed for the entire processes time-series modeling of aero-engines. The TSMRT framework employs Augmented Dickey-Fuller (ADF) test to classify aero-engine operational data into steady state processes and non-steady state processes, and models the two processes separately using time–frequency feature neural network (TFNN) model and MRT model. The time-series forecasting performance of the MRT model is validated on three publicly available benchmark datasets and an ablation study is conducted on turbofan engine test bench datasets. The results indicate that the TSMRT framework demonstrates superior forecasting performance across steady state, non-steady state, and entire processes. This framework stands out as a novel method for component-level modeling of aero-engines.

MIST: Multi-instance selective transformer for histopathological subtype prediction

Accurate histopathological subtype prediction is clinically significant for cancer diagnosis and tumor microenvironment analysis. However, achieving accurate histopathological subtype prediction is a challenging task due to (1) instance-level discrimination of histopathological images, (2) low inter-class and large intra-class variances among histopathological images in their shape and chromatin texture, and (3) heterogeneous feature distribution over different images. In this paper, we formulate subtype prediction as fine-grained representation learning and propose a novel multi-instance selective transformer (MIST) framework, effectively achieving accurate histopathological subtype prediction. The proposed MIST designs an effective selective self-attention mechanism with multi-instance learning (MIL) and vision transformer (ViT) to adaptive identify informative instances for fine-grained representation. Innovatively, the MIST entrusts each instance with different contributions to the bag representation based on its interactions with instances and bags. Specifically, a SiT module with selective multi-head self-attention (S-MSA) is well-designed to identify the representative instances by modeling the instance-to-instance interactions. On the contrary, a MIFD module with the information bottleneck is proposed to learn the discriminative fine-grained representation for histopathological images by modeling instance-to-bag interactions with the selected instances. Substantial experiments on five clinical benchmarks demonstrate that the MIST achieves accurate histopathological subtype prediction and obtains state-of-the-art performance with an accuracy of 0.936. The MIST shows great potential to handle fine-grained medical image analysis, such as histopathological subtype prediction in clinical applications.

STRATEGIC FRAMEWORK OF DIGITAL TRANSFORMATION TOWARDS CLIMATE-NEUTRAL ECONOMY

The article presents a strategic framework for digital transformation aimed at achieving a climate-neutral economy. It proposes digitalization as the primary driver of the green-digital transition, essential for reducing greenhouse gas emissions and optimizing energy use. The study identifies potential barriers and risks associated with this transition and offers strategic dimensions to harmonize digital and green initiatives of global economy. This approach emphasizes the need for international cooperation, the development of global standards, and the importance of bridging the digital divide. The article also highlights the critical role of innovation in key industries to support climate neutrality in both the short and long term. The research underscores the necessity of integrating social, economic, and ethical considerations into the green-digital transition, ensuring that sustainable digitalization contributes to long-term economic growth and environmental sustainability.

Global–Local Deep Fusion: Semantic Integration with Enhanced Transformer in Dual-Branch Networks for Ultra-High Resolution Image Segmentation

The fusion of global contextual information with local cropped block details is crucial for segmenting ultra-high resolution images. In this study, A novel fusion mechanism termed global–local deep fusion (GL-Deep Fusion) is introduced, based on an enhanced transformer architecture that efficiently integrates global contextual information and local details. Specifically, we propose the global–local synthesis networks (GLSNet), a dual-branch network where one branch processes the entire original image, while the other branch handles cropped local patches as input. The feature fusion of different branches in GLSNet is achieved through GL-Deep Fusion, significantly enhancing the accuracy of ultra-high resolution image segmentation. Identifying tiny overlapping items is a task where the model excels, demonstrating its particular effectiveness. To optimize GPU memory utilization, a dual-branch architecture was meticulously designed. This architecture proficiently leverages the features it extracts and seamlessly integrates them into the enhanced transformer framework of GL-Deep Fusion. Benchmarks on the DeepGlobe and Vaihingen datasets demonstrate the efficiency and accuracy of the proposed model. It significantly reduces GPU memory usage by 24.1% on the DeepGlobe dataset, enhancing segmentation accuracy by 0.8% over the baseline model. On the Vaihingen dataset, our model delivers a Mean F1 score of 90.2% and achieves a mIoU of 90.9%, highlighting its exceptional memory efficiency and segmentation precision.

Towards a transformative health humanities approach in teaching the Sustainable Development Goals (SDGs)

The adoption of the United Nations' Sustainable Development Goals (SDGs) marks a significant shift in global political agendas, emphasising sustainability in various fields, including health. To engage meaningfully with sustainability,...

THE DOCTRINE OF SUFI PHILOSOPHY AS A POWERFUL ANTIDOTE TO GLOBAL TERRORISM

The rise of global terrorism presents a critical challenge to peace and stability worldwide, fueled by divisive ideologies that promote hatred and violence. There is a growing need for alternative frameworks that promote understanding, tolerance, and reconciliation. This paper explores the doctrine of Sufi philosophy as a powerful antidote to global terrorism, examining its principles and practices in fostering peace and countering extremist ideologies. A comprehensive review of literature on Sufi philosophy, Islamic mysticism, and counter-terrorism strategies was conducted to elucidate the principles and practices of Sufism in promoting peace and countering extremist narratives. The findings reveal that Sufi philosophy embodies principles of love, compassion, and spiritual enlightenment, potent antidotes to the rhetoric of hatred and violence propagated by extremist groups. Sufi teachings emphasize the interconnectedness of humanity, fostering a culture of tolerance, understanding, and mutual respect. In conclusion, the doctrine of Sufi philosophy offers a transformative framework for countering global terrorism by addressing its root causes of ignorance, fear, and alienation. Through its emphasis on love, compassion, and spiritual enlightenment, Sufism provides a compelling alternative to extremist ideologies, inspiring individuals and communities worldwide to embrace a path of peace, unity, and understanding.

Inferring single-cell spatial gene expression with tissue morphology via explainable deep learning.

The spatial arrangement of cells is vital in developmental processes and organogenesis in multicellular life forms. Deep learning models trained with spatial omics data uncover complex patterns and relationships among cells, genes, and proteins in a high-dimensional space, providing new insights into biological processes and diseases. State-of-the-art in silico spatial multi-cell gene expression methods using histological images of tissue stained with hematoxylin and eosin (H&E) to characterize cellular heterogeneity. These computational techniques offer the advantage of analyzing vast amounts of spatial data in a scalable and automated manner, thereby accelerating scientific discovery and enabling more precise medical diagnostics and treatments. In this work, we developed a vision transformer (ViT) framework to map histological signatures to spatial single-cell transcriptomic signatures, named SPiRiT ( S patial Omics P rediction and R eproducibility integrated T ransformer). Our framework was enhanced by integrating cross validation with model interpretation during hyper-parameter tuning. SPiRiT predicts single-cell spatial gene expression using the matched histopathological image tiles of human breast cancer and whole mouse pup, evaluated by Xenium (10x Genomics) datasets. Furthermore, ViT model interpretation reveals the high-resolution, high attention area (HAR) that the ViT model uses to predict the gene expression, including marker genes for invasive cancer cells ( FASN ), stromal cells ( POSTN ), and lymphocytes ( IL7R ). In an apple-to-apple comparison with the ST-Net Convolutional Neural Network algorithm, SPiRiT improved predictive accuracy by 40% using human breast cancer Visium (10x Genomics) dataset. Cancer biomarker gene prediction and expression level are highly consistent with the tumor region annotation. In summary, our work highlights the feasibility to infer spatial single-cell gene expression using tissue morphology in multiple-species, i.e., human and mouse, and multi-organs, i.e., mouse whole body morphology. Importantly, incorporating model interpretation and vision transformer is expected to serve as a general-purpose framework for spatial transcriptomics.

Identification of the Key Issues and Technical Paths for Intelligent Operation of Water Source Heat Pump Energy Stations Applying Digital Twin Technology

To address the challenges posed by global climate change, developing green energy systems characterized by informatization, digitalization, and intelligence is crucial for achieving carbon neutrality. This article is a research report type paper on water source heat pump (WSHP) energy stations, aiming to use digital twin technology and other information technologies to resolve conflicts between clean energy development and efficient energy utilization. The primary objective of this study is to identify and analyze issues in traditional energy station operations and management systems. Based on this analysis, specific technical solutions are proposed, including pathways for technological research, methodologies, and content. The results provide a comprehensive theoretical framework for the intelligent transformation of energy station systems and essential technical support for the WSHP energy station project in the Hankou Binjiang International Business District. The findings have significant implications for the widespread adoption of WSHP energy stations and the achievement of national carbon neutrality goals.

Transformation or more of the same? The EU’s deforestation-free products regulation through a radical transformation lens

The European Union (EU) is a major consumer of forest-risks commodities and therefore a key contributor to imported deforestation. In 2023, the EU Regulation on Deforestation-Free Products (EUDR) entered into force, which aims to counter the imported deforestation resulting from EU consumption of cattle, cocoa, coffee, palm oil, rubber, soy, wood and derived products. With this regulation, the EU is seemingly recognising its position as one of the world's largest consumers of natural resources and land-consuming agricultural products and adjusting its economy in the face of the global biodiversity, climate and planetary justice crises. In this article, we develop a radical transformation framework to put the EUDR to the test, asking to what extent the EUDR provides a truly transformative response to today’s planetary socio-ecological and justice crises. We develop a four-pronged radical transformation framework bridging insights from transformation, degrowth, Indigenous and decolonial environmental justice, restorative justice and science and technology studies. Concretely, we distil four tenets of radical transformation: onto-epistemic transformation, political transformation, economic transformation and judicial transformation. We apply our framework tentatively to the EUDR to illustrate its practical use, raise questions and highlight possible points of tension. Our preliminary findings point to the simultaneous existence in the EUDR’s design of elements potentially paving the way for counter-hegemonic (re)interpretations and uses on the ground and elements reflecting and reinscribing existing hegemonies.

Advocating integration of human responses in the flood resilience framework for inland cities of northern China

ABSTRACT Frequent urban flooding poses a predominant challenge to resilience managers and the public in inland cities. Nevertheless, various facets of the formation, development, and response to inland flooding in developing countries are still less understood. Using the statistics data of typical flood events in China, this paper examines the causes of devastation in inland areas of the north combined with a comprehensive risk assessment methodology and explores a nuanced understanding of resilience. The conclusions reveal that the inland cities of northern China are facing an imminent threat from escalating flood hazards. 11% of the flood-influenced population and 17% of the disaster-caused economic losses are concentrated in the region. Increased extreme precipitation, heightened exposure and vulnerability, and a sluggish human response to crises are main factors contributing to these complex disasters. The study also proposes a transformative flood resilience framework for inland flooding from three dimensions, emphasizing the integration of human responses. The three aspects of the framework are summarized based on the driving factors mentioned earlier and a simplified multidimensional perspective of resilient cities, reflecting the adaptive and evolutionary features of resilience. Hence, incorporating the resilience strategies to urban flood risk management is a relatively practical approach to urban flood risk management in inland regions.

Dismantling linear lock‐ins in the Australian AEC industry: A pathway to a circular economy

AbstractThe architecture, engineering, and construction (AEC) industry has struggled in its efforts to transition to a circular economy (CE) due to lock‐ins, where the industry remains entrenched in practices that resist this much‐needed transition. The solutions implemented so far have been unsuccessful. This failure can largely be attributed to the prevailing paradigm, which assumes that by sequentially removing barriers transition will naturally follow. However, this assessment overlooks the “lock‐in” effect of practices within the industry's linear model. A comprehensive, holistic “whole of industry” approach is essential to uncover these lock‐ins. In this context, O'Brien's “three transformational spheres framework” is introduced to examine how Australia's AEC sector is deterred from transitioning to a CE by three groups of lock‐ins, and provide viable recommendation to tackle them. Fifteen industry experts were interviewed, covering the three dimensions of the framework. The findings specify three levels of intervention that must be addressed in the industry and recommend an order for tackling them: first, political systems and structures; second, practical behavioral and technical responses; and finally, personal beliefs and values. This article contributes significantly to the field by outlining a comprehensive array of strategies for industry transition. Implementing these strategies in the discussed order has the potential to catalyze the long‐awaited transformation of the industry. This advances the theoretical framework concerning the adoption of a CE within the AEC sector and provides a reliable reference for policymakers, practitioners, and advocates who are orchestrating this transformative journey based on circular principles.

Visual tracking based on spatiotemporal transformer and fusion sequences

Currently, Transformer-based visual tracking methods have exhibited impressive performance. However, despite their widespread adoption, they still have certain limitations. For example, the design of the Transformer framework is somewhat original and redundant, resulting in lower efficiency. In addition, their application methods lack time consideration, and there is a lack of spatiotemporal correlation between tracking video sequences and predicting coordinate sequences, making it difficult to effectively integrate, and the robustness of corresponding tracking templates is insufficient. To address these issues, we propose a new visual tracking method (STFS). Firstly, it introduces a novel Flatten Transformer architecture, which, in comparison to previous modules, offers enhanced efficiency and expressiveness. Secondly, it takes multi frame feature maps and bounding box coordinates as inputs, integrates spatiotemporal information through the spatiotemporal sequence attention module, and provides relevant sequences for historical trend prediction. Finally, it uses diffusion methods to construct tracking templates and improve stability. To verify the performance of the tracker, we conducted experiments on benchmark datasets including GOT-10 K, LaSOT, TrackingNet, VOT2020, OTB100, and UAV123. The results demonstrate that STFS has achieved competitive experimental results.

Broken Buddhas, burials, and sanctuary-adjacent sanctuaries: the ancestral animist archaeologies of Angkor’s ancient places and things

ABSTRACT A growing body of scholarship exploring Cambodia’s cultural-religious environment alongside reinterpretations of ancient Angkorian epigraphy has illuminated the enduring sacredness of Cambodia’s ancient religious places and objects. This assertion comes despite apparent dissociation of these elements from their original ascribed identities (Brahmano-Buddhist) and disuse as focal points of politico-religious congregation at some point in the past. Although documented within Cambodian archaeological studies since the 20th century, fieldwork conducted at ancient Theravāda Buddhist monasteries (vihāra/praḥ vihār) within the Khmer civic-ceremonial center of Angkor Thom between 2017 and 2023 have substantiated that these ancient statues and holy spaces continued to serve as equivalently spiritual, highly localized arenas of ancestral animist practices and cultural-historical negotiation over time. This paper assesses several categories of these archaeological data within the framework of reidentification, reuse, and transformation beyond initial discard, including the deposition of statuary and acts of place-making in the vicinity of older ruins.

A Review of the U.S.-China Science and Technology Cooperation Agreement and Its Prospect–Based on the Perspective of Game Theory

《中美科技合作协定》的签署历程与发展历史解释了中美两国签订该协定的动因与维持双边科技合作关系的考量，并深刻揭示了两国科技合作关系背后的经济与政治博弈。现阶段，美国对华战略认知的调整与其将科技问题安全化的倾向导致了《中美科技合作协定》遭遇续签危机，其未来走向成为中美各界尤其关心的问题。博弈论为中美科技关系的战略转型提供有效的分析框架。其中，中美猎鹿博弈模型揭示了美国对华科技脱钩力度较弱的情况下双方进行合作的进路，而重复囚徒困境博弈模型则指出长期合作与互信的建立才是中美实现双赢的关键。中美两国应努力消除分歧，构建合作与竞争并存的新型关系，避免零和游戏，共同推动科技进步和全球发展。 The signing process and development of the US-China Science and Technology Cooperation Agreement (US-China S&T Cooperation Agreement) explain the reasons for the signing of the agreement and the considerations for maintaining the bilateral S&T cooperation relationship between China and the US, and deeply reveals the economic and political games behind the S&T cooperation relationship between the two countries. At this stage, the adjustment of the U.S. strategic perception of China and its tendency to securitize science and technology issues have led to the renewal crisis of the US-China S&T Cooperation Agreement, and its future direction has become an issue of particular concern to all sectors in the U.S. and China. Game theory provides an effective analytical framework for the strategic transformation of Sino-US S&T relations. The Stag Hunt Game Model reveals the way forward for cooperation between the two sides in the case of weak US S&T decoupling from China, while the Repeated Prisoner’s Dilemma Game Model points out that the establishment of long-term cooperation and mutual trust is the key to achieving a win-win situation for both China and the US. China and the United States should strive to eliminate differences, build a new type of relationship where cooperation and competition coexist, avoid zero-sum games, and jointly promote scientific and technological progress and global development.

High-fidelity level-set modeling of diffusive solid-state phase transformations for polycrystalline materials

The formation of microstructures in metallic alloys during hot metal forming involves simultaneous metallurgical complex phenomena. Traditional high-fidelity numerical frameworks used on the polycrystalline scale tend to focus on single-phase microstructures or isolate phase transformations from grain boundary migration mechanisms. The level-set method is highlighted as effective in proposing a global framework for modeling multiphase polycrystalline materials and diffusive solid-state phase transformations. This framework includes novel techniques for efficient large-scale microstructural representation, strong coupling with ThermoCalc software for real-time thermodynamic data, application for ternary alloys and beyond by taking solute drag aspects, and the use of advanced nucleation models. Numerous applications are then illustrated.

Development and Initial Validation of the Counselor Self-Reflection Inventory for Counselors and Counselors-in-Training

Objective The purpose of this study was to develop and validate the Counselor Self-Reflection Inventory (CSRI) from a Transformative Learning Theory framework for counselors, and counselors-in-training to use in clinical and training settings. Method A sample of 351, mostly female (86.89%), white (85.19%), counselors with MS or MA (88.08%) graduate degrees completed the online survey to assess their self-reflective practice. Results The 39-item CSRI was validated using Confirmatory Factor Analysis. A one factor model of counselor self-reflection suggested indices are acceptable and indicate an overall good model fit (CFI = 0.961, TLI = 0.958, RMSEA = 0.04 (90% CI[0.033; 0.049]), and SRMR = 0.065.) with standardized covariance between the CSRI, self-reflection, and awareness at 0.508. Conclusions Counselors and counselor educators can use the CSRI to understand strengths and areas of growth in their work. Counselors and counselors-in-training can take the CSRI to enhance self-reflective practice that promote positive identity development.

Leveraging Vision Transformers for Detecting Porosity in Wind Energy Composites

The structural robustness and operational efficiency of wind turbine rotor blades are crucial for the overall effectiveness of wind energy systems, often constructed with fiber-reinforced polymers (FRPs) and adhesives. However, porosity within these materials poses a significant threat, weakening structural strength and effectiveness. Air pockets lead to stress concentration points, reducing load-carrying capacity and elevating the risk of blade failure, especially under dynamic wind loads. Manual detection of these air pockets is laborious, necessitating automated inspection techniques. Advanced imaging technologies, such as computed tomography (CT) scanning and deep learning, hold promise for identifying and quantifying porosity in FRPs and adhesives, reducing labor while enhancing accuracy. The study introduces a transformer-based model for porosity detection, departing from convolution-based methods, emphasizing the incorporation of global context throughout the network. Leveraging Vision Transformer (ViT) framework advances, the model is applied to porosity segmentation in wind energy blades, showing promising results with limited datasets. The prospect of using larger datasets suggests potential for a versatile solution in segmenting porosity or voids in various wind energy blade composites, including adhesives.

Spach Transformer: Spatial and Channel-Wise Transformer Based on Local and Global Self-Attentions for PET Image Denoising.

Position emission tomography (PET) is widely used in clinics and research due to its quantitative merits and high sensitivity, but suffers from low signal-to-noise ratio (SNR). Recently convolutional neural networks (CNNs) have been widely used to improve PET image quality. Though successful and efficient in local feature extraction, CNN cannot capture long-range dependencies well due to its limited receptive field. Global multi-head self-attention (MSA) is a popular approach to capture long-range information. However, the calculation of global MSA for 3D images has high computational costs. In this work, we proposed an efficient spatial and channel-wise encoder-decoder transformer, Spach Transformer, that can leverage spatial and channel information based on local and global MSAs. Experiments based on datasets of different PET tracers, i.e., 18F-FDG, 18F-ACBC, 18F-DCFPyL, and 68Ga-DOTATATE, were conducted to evaluate the proposed framework. Quantitative results show that the proposed Spach Transformer framework outperforms state-of-the-art deep learning architectures.

THE SETAs AND S-SYSTEM SKILLS DEVELOPMENT IN SOUTH AFRICA AND BRAZIL: A TRANSFORMATIVE SOCIAL POLICY PERSPECTIVE

Relevance. The African Union declared 2024 the Year of Education, Educating an African Fit for the 21st Century and the Agenda 2063 aspiration seeks to catalyse an education and skills revolution for Africa’s industrialisation. This places Technical and Vocational Education and Training at the centre in skills development on the continent. As part of their endogenous initiatives at improving skills, knowledge and abilities desired for work and enhance industrial competitiveness South Africa and Brazil instituted the SETAs system and the S-System, respectively aimed at skills formation and human capital development. Aim. The two BRICS countries of South Africa and Brazil, through Acts of Parliament, institutionalised skills development programs which had seen them emerge as leading economies on their respective continents. What key lessons can be drawn for other African countries in their quest for industrialisation in the AU Year of the Year of Education to foster the relevant and competitive skills needed for the 21st century. Methods. The case studies of the S-system of Brazil and the SETAs system of South Africa are presented using a case study approach. Examining the SETAs and S-systems using a Transformative Social Policy framework, the paper adopts a case study approach seeks to answer the following questions: (1) what kind of institutional linkages and coordination between industry/business and skills/education training institutions need to be in place for adequate and appropriate human capital formation in Africa? (2) what are the appropriate funding mechanisms for human capital development suitable for African countries to adopt? (3) What coordination and facilitating role should be adopted by states in African countries to facilitate industry/training institutional linkages? Results and conclusions. Key finding indicates that while the SETAs in South Africa are widely recognised by employers and give holders a big advantage in securing employment, TVET-market coordination, regional coordination for area-specific industrialisation and appropriate skills funding mechanism are key for competitive skills development and provide practical and policy lessons for African governments.