Technical Background Research Articles

In-context learning enables multimodal large language models to classify cancer pathology images

Medical image classification requires labeled, task-specific datasets which are used to train deep learning networks de novo, or to fine-tune foundation models. However, this process is computationally and technically demanding. In language processing, in-context learning provides an alternative, where models learn from within prompts, bypassing the need for parameter updates. Yet, in-context learning remains underexplored in medical image analysis. Here, we systematically evaluate the model Generative Pretrained Transformer 4 with Vision capabilities (GPT-4V) on cancer image processing with in-context learning on three cancer histopathology tasks of high importance: Classification of tissue subtypes in colorectal cancer, colon polyp subtyping and breast tumor detection in lymph node sections. Our results show that in-context learning is sufficient to match or even outperform specialized neural networks trained for particular tasks, while only requiring a minimal number of samples. In summary, this study demonstrates that large vision language models trained on non-domain specific data can be applied out-of-the box to solve medical image-processing tasks in histopathology. This democratizes access of generalist AI models to medical experts without technical background especially for areas where annotated data is scarce.

Analyzing Occupational Accidents and Exoskeleton Potential in the Construction Industry in Québec, Canada

Occupational applications Our analysis of construction firm accident data in Québec, Canada, spanning from January 2019 to June 2022, provides critical insights for ergonomics and human factors practitioners. The predominant accidents involved being struck by objects (31%) and falls (15%), with lacerations and bruises comprising 37% of injuries. Notably, carpenters had the highest accident rate (22%), prompting attention to task-specific safety measures. We also examined musculoskeletal injuries (MSI), finding that bodily reactions (46%) and overexertion (44%) were primary causes. Importantly, we qualitatively explored the potential of exoskeletons as possible proactive safety measures; our results suggested that in 50% of MSI cases, exoskeletons might have helped to mitigate or eliminate risks. These findings underscore the potential for exoskeletons to enhance safety and productivity in the construction industry, offering opportunities for intervention and preventive measures in ergonomics practice. TECHNICAL ABSTRACT Background: The construction industry is a hazardous working environment, having a relatively high risk of accidents and injuries compared to other industries. Purpose: We aimed to describe the characteristics of work accidents in a large construction firm in Québec, Canada, using accident data from the Health, Safety, and Environment Department. Methods: Our dataset spanned from January 2019 to June 2022 and included 2065 complete entries for analysis. Accidents were categorized using a standardized classification scheme, augmented with additional accident types to provide more precision. Results: The most common type of accident was being struck by an object, comprising 31% of reported incidents, followed by falls at 15%. Lacerations and bruises were the most prevalent injuries, accounting for 37% of cases, with injuries to the hand (31%) and the head/eye (24%) being the most frequent. Among the trades analyzed, carpenters had the highest accident rate at 22%, closely followed by laborers at 20%. We also focused on occupational musculoskeletal injuries (MSI) in the dataset to qualitatively investigate the viability of exoskeletons as a proactive safety measure. Of the 268 incidents categorized as MSI-related (13% of the 2065 events), bodily reactions (46%) and overexertion (44%) were the leading causes. A detailed qualitative analysis of the event descriptions suggested that if exoskeletons had been made available and used, they could have contributed to reducing or eliminating MSI risk in 50% of cases. Conclusions: These results contribute to enhancing safety and productivity in the construction industry by providing insights into work accidents and task characteristics that can be used to improve exoskeleton design and compatibility with the work to be performed.

Advancements in Using AI for Dietary Assessment Based on Food Images: Scoping Review.

To accurately capture an individual's food intake, dietitians are often required to ask clients about their food frequencies and portions, and they have to rely on the client's memory, which can be burdensome. While taking food photos alongside food records can alleviate user burden and reduce errors in self-reporting, this method still requires trained staff to translate food photos into dietary intake data. Image-assisted dietary assessment (IADA) is an innovative approach that uses computer algorithms to mimic human performance in estimating dietary information from food images. This field has seen continuous improvement through advancements in computer science, particularly in artificial intelligence (AI). However, the technical nature of this field can make it challenging for those without a technical background to understand it completely. This review aims to fill the gap by providing a current overview of AI's integration into dietary assessment using food images. The content is organized chronologically and presented in an accessible manner for those unfamiliar with AI terminology. In addition, we discuss the systems' strengths and weaknesses and propose enhancements to improve IADA's accuracy and adoption in the nutrition community. This scoping review used PubMed and Google Scholar databases to identify relevant studies. The review focused on computational techniques used in IADA, specifically AI models, devices, and sensors, or digital methods for food recognition and food volume estimation published between 2008 and 2021. A total of 522 articles were initially identified. On the basis of a rigorous selection process, 84 (16.1%) articles were ultimately included in this review. The selected articles reveal that early systems, developed before 2015, relied on handcrafted machine learning algorithms to manage traditional sequential processes, such as segmentation, food identification, portion estimation, and nutrient calculations. Since 2015, these handcrafted algorithms have been largely replaced by deep learning algorithms for handling the same tasks. More recently, the traditional sequential process has been superseded by advanced algorithms, including multitask convolutional neural networks and generative adversarial networks. Most of the systems were validated for macronutrient and energy estimation, while only a few were capable of estimating micronutrients, such as sodium. Notably, significant advancements have been made in the field of IADA, with efforts focused on replicating humanlike performance. This review highlights the progress made by IADA, particularly in the areas of food identification and portion estimation. Advancements in AI techniques have shown great potential to improve the accuracy and efficiency of this field. However, it is crucial to involve dietitians and nutritionists in the development of these systems to ensure they meet the requirements and trust of professionals in the field.

An ASIC-based system-in-package MEMS gas sensor with impedance spectroscopy readout and AI-enabled identification capabilities

The main advantages of micromachined metal oxide semiconductor (MOX) gas sensors, in particular their low cost, high sensitivity, miniaturization and low power consumption, are compromised by lack of selectivity and long-term sensor signal drifts, which often hinder their application in real-world sensing applications. To increase the sensor data dimensionality, temperature cycled operation (TCO) and impedance spectroscopy (IS) readout was proposed, enabling the acquisition of large datasets from a single sensor, as suitable to be exploited by machine learning techniques. Furthermore, recent studies have shown how the IS readout minimizes the issues represented by the sensor long-term signal drift.The aim of this work is the development of a smart system-in-package, integrating a state-of-the-art micromachined MOX gas sensor based on a wafer-level nanostructured sensing layer with a novel versatile sensor interface chip, capable of high performance IS readout. We propose a miniaturized sensing device, which is a small, lightweight low power consumption chemical sensing micro-system with fast and accurate classification capabilities, suitable to be deployed in long multi-node sensing cables or on unmanned aerial vehicles. A software suite comprising a sensor training wizard was also developed, enabling users with no technical background to train the sensing system.

IgE immunoadsorption: technical background, functionality, and first clinical experience

Summary Background The prevalence of allergic diseases has risen in the 21st century, drawing attention to specific therapeutic and preventive strategies. Due to the key role of immunoglobulin E (IgE) in the development of allergic reactions, IgE represents a key target treatment. In this scenario, IgE immunoadsorption (IgE-IA) has been investigated as a procedure that selectively removes circulating IgE antibodies from the bloodstream of patients with atopy. Methods This narrative review aims to critically summarize the current insights regarding IgE-IA in the context of the management of allergic diseases, ranging from the rationale to the technical aspects, as well as the benefits and unmet needs. Results IgE-IA might be a treatment strategy in well-selected patients with allergic diseases. IgE depletion through sessions of IgE-IA results in immediate clinical improvement and might be useful in acute situations when a rapid clinical response is required or when classic approaches are contraindicated or ineffective. Due to the reduced effectiveness over time, IgE-IA could be a valid first approach before starting another IgE depletion therapy, such as omalizumab, when its commencement would otherwise be contraindicated by too-high serum IgE levels. Conclusion Overall, IgE-IA is safe and well tolerated; however, this procedure is currently difficult to implement in routine clinical practice because of costs, time demands, need for hospitalization, and the invasiveness of the procedure, with the associated risks related to the necessity of venous catheterization.

Elastography as a Discriminator Between Fibrotic and Inflammatory Strictures in Crohn's Disease: A Dead End or Bright Future in Clinical Decision-Making? Critical Review.

Crohn's disease (CD) is a complex systemic entity, characterized by the progressive and relapsing inflammatory involvement of any part of the gastrointestinal tract. Its clinical pattern may be categorized as penetrating, stricturing or non-penetrating non-stricturing. In this paper, we performed a database search (Pubmed, MEDLINE, Mendeley) using combinations of the queries "crohn", "stricture" and "elastography" up to 19 June 2024 to summarize current knowledge regarding the diagnostic utility of ultrasound (US) and magnetic resonance (MR) elastography techniques in the evaluation of stricturing CD by means of an assessment of the transmural intestinal fibrosis. We decided to include papers published since 1 January 2017 for further evaluation (n = 24). Despite growing collective and original data regarding numerous applications of mostly ultrasound elastography (quantification of fibrosis, distinguishing inflammatory from predominantly fibrotic strictures, assessment of treatment response, predicting disease progression) constantly emerging, to date, we are still lacking a uniformization in both cut-off values and principles of measurements, i.e., reference tissue in strain elastography (mesenteric fat, abdominal muscles, unaffected bowel segment), units, not to mention subtle differences in technical background of SWE techniques utilized by different vendors. All these factors imply that ultrasound elastography techniques are hardly translatable throughout different medical centers and practitioners, largely depending on the local experience. Nonetheless, the existing medical evidence is promising, especially in terms of possible longitudinal comparative studies (follow-up) of patients in the course of the disease, which seems to be of particular interest in children (lack of radiation, less invasive contrast media) and terminal ileal disease (easily accessible).

Brief analysis of the existing scientific and technical background in the development of military exoskeletons

The article provides a brief analysis of the existing scientific and technical groundwork created by Russian and for-eign experts in the field of development of military exoskeletons (exoskeletal structures). Based on the results of the analysis, the main trends and further prospects for the development of military exoskeletons have been identified.

Private entrepreneurs’ technical background and innovation output: The perspective of innovation quantity and quality

This study investigates the impact of entrepreneurs' technical backgrounds on innovation output from the perspective of innovation quantity and quality, using data from A-share listed private companies in China between 2009 and 2018. Our analysis reveals that CEOs with technical expertise significantly enhance both the quantity and quality of firm innovation. We identify three key mechanisms through which technical backgrounds promote innovation: increased management attention to innovation, reduced information asymmetry, and enhanced long-term investment. These findings contribute to the literature on Upper Echelons Theory and corporate innovation, offering practical insights for private enterprises aiming to achieve talent-driven technological innovation and foster high-quality development.

What factors distinguish overlapping Data job postings? Towards ML-based models for job category’s factors prediction

Distinguishing between roles like Data Scientist, Data Engineer, Data Analyst, and Business Intelligence Developer can be challenging, as there can be overlap in responsibilities, focuses, and skill sets across these positions. By understanding these distinctions, job seekers can better align their skills and interests with the specific requirements and factors of each role, thereby increasing their chances of finding a fulfilling career in the data field. To address what factors distinguish these positions, we developed machine learning models capable of clarifying the distinctions among these positions based on relevant features extracted from the dataset. The proposed learning models leverage relevant features extracted from the dataset to differentiate between roles accurately. Factors such as technical skills, programming languages, educational background, work experience, and certifications likely play crucial roles in distinguishing between these positions. By incorporating these features into the models, they can effectively identify patterns and characteristics unique to each role. The high accuracy (approximately 99%) achieved by these models not only validates their effectiveness but also underscores the importance of understanding the nuances and specific requirements of each role within the data field. Armed with this knowledge, both job seekers and employers can make more informed decisions when it comes to hiring, career planning, and talent acquisition.

Explainable Generative AI (GenXAI): a survey, conceptualization, and research agenda

Generative AI (GenAI) represents a shift from AI’s ability to “recognize” to its ability to “generate” solutions for a wide range of tasks. As generated solutions and applications grow more complex and multi-faceted, new needs, objectives, and possibilities for explainability (XAI) have emerged. This work elaborates on why XAI has gained importance with the rise of GenAI and the challenges it poses for explainability research. We also highlight new and emerging criteria that explanations should meet, such as verifiability, interactivity, security, and cost considerations. To achieve this, we focus on surveying existing literature. Additionally, we provide a taxonomy of relevant dimensions to better characterize existing XAI mechanisms and methods for GenAI. We explore various approaches to ensure XAI, ranging from training data to prompting. Our paper provides a concise technical background of GenAI for non-technical readers, focusing on text and images to help them understand new or adapted XAI techniques for GenAI. However, due to the extensive body of work on GenAI, we chose not to delve into detailed aspects of XAI related to the evaluation and usage of explanations. Consequently, the manuscript appeals to both technical experts and professionals from other fields, such as social scientists and information systems researchers. Our research roadmap outlines over ten directions for future investigation.

Digital Transformation and Management Earnings Forecast

This paper conducts an empirical examination using listed companies on China's Shanghai and Shenzhen A-share markets from 2010 to 2020. The findings indicate that digital transformation has a significant positive effect on the quality of management's earnings forecast. Mechanism analysis reveals that this effect is more pronounced in companies with lower internal control indices, higher levels of corporate governance, and higher agency costs. Digital transformation enhances the quality of corporate information disclosure and mitigates information asymmetry, further improving management's earnings forecast level. Through heterogeneity analysis, further research finds that companies lacking a technical background among senior executives, those in less competitive industries, those with poor growth, and non-state-owned enterprises, experience a more significant positive impact on the quality of management earnings forecasts from digital transformation. The conclusions provide new evidence for continuing to advance digital transformation and elucidate the benign cycle mechanism between the capital market and the national economy. This study also proposes new perspectives for the regulation of predictive information disclosure by enterprises.

Natural disasters detection using explainable deep learning

Deep learning applications have far-reaching implications in people’s daily lives. Disaster management professionals are becoming increasingly interested in applying deep learning to prepare for and respond to natural disasters. In this paper, we aim to assist natural disaster management professionals in preparing for disasters by developing a framework that can accurately classify natural disasters and interpret the results using a combination of a deep learning model and an XAI method to ensure reliability and ease of interpretation without a technical background. Two main aspects categorize the novelty of our work. The first is utilizing pre-trained Models such as VGGNet19, ResNet50, and ViT for accurate classification of natural disaster images. The second is implementing three explainable AI techniques-Gradient-weighted Class Activation Mapping (Grad-CAM), Grad CAM++, and Local Interpretable Model-agnostic Explanations (LIME) to ensure the interpretability of the model’s predictions, making the decision-making process transparent and reliable. Experiments on the Natural disaster datasets (Niloy et al. 2021) and MEDIC with a ViT-B-32 model achieved a high accuracy of 95.23%. Additionally, explainable artificial intelligence techniques such as LIME, Grad-CAM, and Grad-CAM++ are used to evaluate model performance and visualize decision-making. Our code is available at.11https://github.com/tariqshaban/disaster-classification-with-xai.

Research on the Influence of CEO's Technical Background Characteristics on the Digital Transformation Effectiveness of Manufacturing Enterprises: Taking the Marketization Process as the Moderating Variable

Research on the Influence of CEO's Technical Background Characteristics on the Digital Transformation Effectiveness of Manufacturing Enterprises: Taking the Marketization Process as the Moderating Variable

Euclid. III. The NISP instrument

The Near-Infrared Spectrometer and Photometer (NISP) on board the satellite provides multiband photometry and $R slitless grism spectroscopy in the 950--2020\,nm wavelength range. In this reference article, we illuminate the background of NISP's functional and calibration requirements, describe the instrument's integral components, and provide all its key properties. We also sketch the processes needed to understand how NISP operates and is calibrated as well as its technical potentials and limitations. Links to articles providing more details and the technical background are included. The NISP's 16 H2RG detectors with a plate scale of $ deliver a field of view of 0.57\,deg$^2$. In photometric mode, NISP reaches a limiting magnitude of sim \,24.5\,AB\,mag in three photometric exposures of about 100\,s in exposure time for point sources and with a S/N of five. For spectroscopy, NISP's point-source sensitivity is a SNR = 3.5 detection of an emission line with flux sim \,$2 $ integrated over two resolution elements of 13.4\ in 3times 560\,s grism exposures at 1.6\ (redshifted Halpha ). Our calibration includes on-ground and in-flight characterisation and monitoring of the pixel-based detector baseline, dark current, non-linearity, and sensitivity to guarantee a relative photometric accuracy better than 1.5<!PCT!> and a relative spectrophotometry better than 0.7<!PCT!>. The wavelength calibration must be accurate to 5\ or better. The NISP is the state-of-the-art instrument in the near-infrared for all science beyond small areas available from HST and JWST -- and it represents an enormous advance from any existing instrumentation due to its combination of field size and high throughput of telescope and instrument. During six-year survey covering 14\,000\,deg$^2$ of extragalactic sky, NISP will be the backbone in determining distances of more than a billion galaxies. Its near-infrared data will become a rich reference imaging and spectroscopy data set for the coming decades.

Does Scaling Up Pig Farming Promote Carbon Neutrality Among Pig Farmers?

Pig farming is a major source of carbon emissions in China's livestock industry, and the promotion and application of carbon neutral technologies dedicated to carbon reduction and sequestration is a key measure to promote the realization of carbon neutrality in pig farming industry. In the absence of a system for carbon emission reduction in agriculture, it is important to explore whether large-scale pig farming can spontaneously promote the application of carbon neutral technologies by pig farmers. Combined with the technical background of carbon neutrality in pig farming industry and using the field survey data of 468 pig farmers in Shandong Province, this paper empirically analyzed the effect of large-scale pig farming on the carbon neutral behavior of pig farmers. The findings are threefold. Firstly, because of the effect of economies of scale in the application of carbon neutral technologies, large-scale pig farming can promote the carbon neutral behavior of pig farmers. However, the effect of economies of scale in the application of carbon neutral technologies has a boundary, and the annual output of pigs should be maintained at 3000 to 4999. Secondly, among the effects of pig breeding scale on the application and its degree of composting project, state of planting and breeding cycle has no regulating effect. Thirdly, among the effects of pig breeding scale on the application and its degree of biogas engineering, both state of planting and breeding cycle and subsidy for construction of biogas digester have negative regulating effects. In this paper, the pig farming industry is taken as a typical example to reveal the endogenous incentive mechanism of promoting agricultural carbon emission reduction, which provides theoretical reference and experience enlightenment for the formulation of effective policies to promote agricultural carbon neutrality, which are different from industrial industries.

Counterfactual Explanation Analytics: Empowering Lay Users to Take Action Against Consequential Automated Decisions

Machine learning is routinely used to automate consequential decisions about users in domains such as finance and healthcare, raising concerns of transparency and recourse for negative outcomes. Existing Explainable AI techniques generate a static counterfactual point explanation which recommends changes to a user's instance to obtain a positive outcome. Unfortunately, these recommendations are often difficult or impossible for users to realistically enact. To overcome this, we present FACET, the first interactive robust explanation system which generates personalized counterfactual region explanations. FACET's expressive explanation analytics empower users to explore and compare multiple counterfactual options and develop a personalized actionable plan for obtaining their desired outcome. Visitors to the demonstration will interact with FACET via a new web dashboard for explanations of a loan approval scenario. In doing so, visitors will experience how lay users can easily leverage powerful explanation analytics through visual interactions and displays without the need for a strong technical background.

Case study of automotive NVH topics induced by HVAC systems integration

Within the automotive market, HVAC systems are of critical importance in ensuring passengers' thermal comfort inside the car cabin as well as complying with regulatory defogging functions. These systems involve various components such as blowers, thermal exchangers, actuators... over a wide range of well-known technologies and also innovations on newly introduced designs and products. Currently, within the established electrification trend worldwide, HVAC systems became one of the most important embedded systems, possibly inducing major contributions of noise and vibration. These NVH topics can emerge through various transfer paths inside the car cabin causing significant discomfort to passengers. During development and production stages, the NVH behavior at component level is mastered by both suppliers and OEMS according to internal standards and specifications. However, implementation into the complete vehicle can induce a specific NVH behavior unrelated to the compliance status at component level. Integration impact cases are highly reinforced within electrification trends as they involve more and more design changes with less technical background. Through a practical case study, this paper focuses on the NVH behavior induced by the integration of HVAC systems in the US market. Transfer paths analysis and solutions are investigated to improve the NVH behavior and set the product intrinsic limits giving the targeted vehicle.

A decision-making framework for determinants of an organisation’s readiness for smart warehouse

Demanding, competitive business dynamics drive organisations to continually evaluate and adopt automation in their structural setups, particularly warehouses, for growth and profitability. Industry 4.0 technologies are revolutionising warehouse operations, resulting in cost savings, increased efficiency, and a more environmentally friendly approach for businesses. The readiness of organisations to adopt these technologies is evaluated with the help of hard and soft determinants selected through a theoretical framework of Technology–Organisation-Environment (TOE). In this study, a unique method is used that combines the Modified Delphi method, content validity for scale validation, expert opinion consensus tests for agreement, ranking determinants using the Best Worst Method (BWM), and rank validation through sensitivity analysis for warehousing industry experts and organisations. Warehouse service providers who convert their warehouses into smart modes to serve fast-moving consumer goods companies in Tier-2 cities are chosen for the structural framework. Results present the role of the ‘operational efficiency of technology’, ‘top management’, ‘employee’s education’, and ‘technical background’ as the most effective driving determinants. This research provides valuable information to researchers and managers trying to upgrade their warehouse into a smart warehouse.

SAARC super smart grid: Navigating the future - unleashing the power of an energy-efficient integration of renewable energy resources in the saarc region

The major problems in the power system network of the South Asian Association for Regional Cooperation (SAARC) region include the utilization of substantial fossil fuels, which leads to significant emissions of CO2. Moreover, due to the depletion of fossil fuels, it is very difficult to manage the demand requirement of the SAARC regions, which leads to significant power shortage issues. These problems would be resolved through the integration of the substantial amount of renewable energy resources (RERs) that are available in excess in the SAARC regions into the power system networks. This will provide a futuristic roadmap to a sustainable and green energy environment in the SAARC region. In this regard, the SAARC region has devised a new plan to transform its smart grid infrastructure into the super smart grid (SSG), which would connect the SAARC countries using the high voltage transmission lines utilizing each country's substantial RERs. To accelerate this ecofriendly effort, the key objective of this review study is to evaluate the potential of the SAARC SSG for demand management and emissions reduction. This study provides a comprehensive overview of SSG technology, covering its technical background, management, power flow improvement techniques, power stability techniques, simulation models, and future steps for development in the SAARC region. First, the regional conflicts, energy deficiency, and potential of the SAARC region have been discussed. Second, a potential architecture of the SAARC SSG has been simulated. Third, a novel Markov chain modeling has been presented for demand management in the SAARC regions. The results and discussions section presents findings from simulations and models. The paper also discusses CO2 reduction techniques, including methods for capturing, storing, and utilizing carbon in biofuel. The study also discusses futuristic optimization techniques in SSG, including blockchain, federated learning, reinforcement learning, the metaverse, digital twin technology, artificial intelligence, and cloud computing. The SSG challenges and limitations section summarizes the challenges and limitations faced by SSG. Finally, a roadmap to a sustainable and green energy environment for the SAARC power grids has been identified to support the implementation of the futuristic SAARC SSG. The results exhibited that there is a good potential for successful demand management and emissions reduction through the SAARC SSG.

Competency Mapping of Key Stakeholders in Procurement and Contracting of the National Tuberculosis Elimination Program Goods and Services in an Eastern Indian State

Abstract Background: The National Tuberculosis Elimination Program is a complex program, and the program officers are involved in the procurement and contracting of goods and services under it. A competency mapping was conducted to highlight the current capacity gaps in contract management at state and district levels under the program. Materials and Methods: A descriptive cross-sectional design was adopted with 15 participants involved in procurement and contracting in two districts of the state using a semi-structured questionnaire, and results were presented using the descriptive statistical analysis. Results: A total of 15 different stakeholders from the National Health Mission, State and District Tuberculosis (TB) Cell, State TB Training and Demonstration Centre, and State Medical Corporation participated in the study. The participants had an average of 10.3 years of experience in contracting, and 20% of them had received some training on contracting. The capacity of these respondents varied based on their involvement, previous training (s) received, and technical background. It was further observed that most of the respondents suggested an in-person physical training instead of virtual training. Most of the respondents requested to issue training certificate. Some of the topics that the participants requested for further training include Digital Signature Certification, Government e-Marketplace, General Financial Rule, Public Financial Management System, monitoring of performance indicators, and bid evaluation techniques. Moreover, 92% of the respondents felt that there is a training need on procurement and contracting. Conclusions: Many a time, the program officers come from a technical background not relevant to the contracting and procurement of goods and services and require regular on-the-job support and training to aid in effective program delivery.