Search results

The purpose of this study is to investigate and demonstrate the advancements achieved in the field of chest X-ray image captioning through the utilization of dynamic convolutional encoder–decoder networks (DyCNN). Typical convolutional neural networks (CNNs) are unable to capture both local and global contextual information effectively and apply a uniform operation to all pixels in an image. To address this, we propose an innovative approach that integrates a dynamic convolution operation at the encoder stage, improving image encoding quality and disease detection. In addition, a decoder based on the gated recurrent unit (GRU) is used for language modeling, and an attention network is incorporated to enhance consistency. This novel combination allows for improved feature extraction, mimicking the expertise of radiologists by selectively focusing on important areas and producing coherent captions with valuable clinical information.

In this study, we have presented a new report generation approach that utilizes dynamic convolution applied Resnet-101 (DyCNN) as an encoder (Verelst and Tuytelaars, 2019) and GRU as a decoder (Dey and Salemt, 2017; Pan et al., 2020), along with an attention network (see Figure 1). This integration innovatively extends the capabilities of image encoding and sequential caption generation, representing a shift from conventional CNN architectures. With its ability to dynamically adapt receptive fields, the DyCNN excels at capturing features of varying scales within the CXR images. This dynamic adaptability significantly enhances the granularity of feature extraction, enabling precise representation of localized abnormalities and structural intricacies. By incorporating this flexibility into the encoding process, our model can distil meaningful and contextually rich features from the radiographic data. While the attention mechanism enables the model to selectively focus on different regions of the image during caption generation. The attention mechanism enhances the report generation process by allowing the model to assign different importance weights to different regions of the image, mimicking human perception. In parallel, the GRU-based decoder adds a critical dimension to the process by ensuring a smooth, sequential generation of captions.

The findings of this study highlight the significant advancements achieved in chest X-ray image captioning through the utilization of dynamic convolutional encoder–decoder networks (DyCNN). Experiments conducted using the IU-Chest X-ray datasets showed that the proposed model outperformed other state-of-the-art approaches. The model achieved notable scores, including a BLEU_1 score of 0.591, a BLEU_2 score of 0.347, a BLEU_3 score of 0.277 and a BLEU_4 score of 0.155. These results highlight the efficiency and efficacy of the model in producing precise radiology reports, enhancing image interpretation and clinical decision-making.

This work is the first of its kind, which employs DyCNN as an encoder to extract features from CXR images. In addition, GRU as the decoder for language modeling was utilized and the attention mechanisms into the model architecture were incorporated.

Explainable artificial intelligence (XAI) has importance in several industrial applications. The study aims to provide a comparison of two important methods used for explainable AI algorithms.

In this study multiple criteria has been used to compare between explainable Ranked Area Integrals (xRAI) and integrated gradient (IG) methods for the explainability of AI algorithms, based on a multimethod phase-wise analysis research design.

The theoretical part includes the comparison of frameworks of two methods. In contrast, the methods have been compared across five dimensions like functional, operational, usability, safety and validation, from a practical point of view.

A comparison has been made by combining criteria from theoretical and practical points of view, which demonstrates tradeoffs in terms of choices for the user.

Our results show that the xRAI method performs better from a theoretical point of view. However, the IG method shows a good result with both model accuracy and prediction quality.

Diagnosing brain tumors is a process that demands a significant amount of time and is heavily dependent on the proficiency and accumulated knowledge of radiologists. Over the traditional methods, deep learning approaches have gained popularity in automating the diagnosis of brain tumors, offering the potential for more accurate and efficient results. Notably, attention-based models have emerged as an advanced, dynamically refining and amplifying model feature to further elevate diagnostic capabilities. However, the specific impact of using channel, spatial or combined attention methods of the convolutional block attention module (CBAM) for brain tumor classification has not been fully investigated.

To selectively emphasize relevant features while suppressing noise, ResNet50 coupled with the CBAM (ResNet50-CBAM) was used for the classification of brain tumors in this research.

The ResNet50-CBAM outperformed existing deep learning classification methods like convolutional neural network (CNN), ResNet-CBAM achieved a superior performance of 99.43%, 99.01%, 98.7% and 99.25% in accuracy, recall, precision and AUC, respectively, when compared to the existing classification methods using the same dataset.

Since ResNet-CBAM fusion can capture the spatial context while enhancing feature representation, it can be integrated into the brain classification software platforms for physicians toward enhanced clinical decision-making and improved brain tumor classification.

This research has not been published anywhere else.

Meeting the patients' requirements as customers of the health care sector is crucially important as a social responsibility. According to the resource constraints, only an efficient utilisation of health services can provide that purpose. This study aims to develop a quantitative assessment framework for radiology centres as a vital section in healthcare to translate the patients' requirements into service quality specifications. This would help to achieve quality improvement by emphasising the voice of customers.

A literature review is conducted to specify the service quality criteria and the patients' requirements related to healthcare and hospitals. Based on the experts' opinions, these criteria and requirements are later customised for the radiology centres. Moreover, the requirements are categorised into five dimensions of SERVQUAL. The interrelations between service elements are also determined through expert group consensus using Pearson correlation. Afterwards, by applying the QFD method, the relations between the requirements and criteria are explored. Additionally, a customer satisfaction survey is executed in Tehran public hospitals to prioritise these requirements and provide an importance-satisfaction analysis.

Based on the result of the case study, service elements are prioritised for improvement, and practical suggestions are provided using the Delphi technique for quality improvement. In addition, a cause-and-effect diagram is presented to highlight the improvement area and provide enhancement suggestions.

This study is the first empirical attempt to benefit from the VOC in evaluating and enhancing the quality of service delivered to radiology patients. In doing so, the study applies a hybrid approach of QFD and SERVQUAL as well as other tools to highlight the improvement area and provide enhancement suggestions. The findings can be readily used by the practitioners.

The applications of Artificial Intelligence (AI) in various areas of professional and knowledge work are growing. Emotions play an important role in how users incorporate a technology into their work practices. The current study draws on work in the areas of AI-powered technologies adaptation, emotions, and the future of work, to investigate how knowledge workers feel about adopting AI in their work.

We gathered 107,111 tweets about the new AI programmer, GitHub Copilot, launched by GitHub and analysed the data in three stages. First, after cleaning and filtering the data, we applied the topic modelling method to analyse 16,130 tweets posted by 10,301 software programmers to identify the emotions they expressed. Then, we analysed the outcome topics qualitatively to understand the stimulus characteristics driving those emotions. Finally, we analysed a sample of tweets to explore how emotional responses changed over time.

We found six categories of emotions among software programmers: challenge, achievement, loss, deterrence, scepticism, and apathy. In addition, we found these emotions were driven by four stimulus characteristics: AI development, AI functionality, identity work, and AI engagement. We also examined the change in emotions over time. The results indicate that negative emotions changed to more positive emotions once software programmers redirected their attention to the AI programmer's capabilities and functionalities, and related that to their identity work.

Overall, as organisations start adopting AI-powered technologies in their software development practices, our research offers practical guidance to managers by identifying factors that can change negative emotions to positive emotions.

Our study makes a timely contribution to the discussions on AI and the future of work through the lens of emotions. In contrast to nascent discussions on the role of AI in high-skilled jobs that show knowledge workers' general ambivalence towards AI, we find knowledge workers show more positive emotions over time and as they engage more with AI. In addition, this study unveils the role of professional identity in leading to more positive emotions towards AI, as knowledge workers view such technology as a means of expanding their identity rather than as a threat to it.

Customers today expect businesses to cater to their individual needs by tailoring the products they purchase to their own preferences. The term “Industry 5.0” refers to a new wave of manufacturing that aims to meet each customer's unique demands. Even while Industry 4.0 allowed for mass customization, that wasn't good enough before, customers today demand individualized products at scale, and Industry 5.0 is driving the transition from mass customization to mass personalization to meet these demands. It caters to the individual needs of each consumer by meeting their demands. More specialized components for use in medicine are made possible by the widespread customization made possible by Industry 5.0. These individualized parts are included into the medical care of the patient to meet their specific needs and preferences. In the current medical revolution, an enabling technology of Industry 5.0 can produce medical implants, artificial organs, bodily fluids, and transplants with pinpoint accuracy. With the advent of AI-enabled sensors, we now live in a world where data can be swiftly analyzed. Machines may be programmed to make complex choices on the fly. In the medical field, these innovations allow for exact measurement and monitoring of human body variables according to the individual's needs. They aid in monitoring the body's response to training for peak performance. It allows for the digital dissemination of accurate healthcare data networks. In order to collect and exchange relevant patient data, every equipment is online.

Digital healthcare once again emerges due to pandemic (Covid-19). Digital healthcare can be minimising the issue of accessibility, availability, accuracy and affordability of healthcare service during a pandemic. Digital healthcare playsa significant role to provide healthcare equity during the pandemic. This article presents the current trends and scenario of digital healthcare with a focus on health equity. The main objective of this chapter is to review the four aces of health equity in the digital healthcare literature. The scope and challenges faced by the policymakers to implementation of digital healthcare to improve health equity. This chapter considers the hybrid literature review based on the bibliometric and the systematic literature based on the various theme, sub-theme, concept and context-related health equity through digital healthcare. This study provides the previous and current research trends and preposition for the future researcher, healthcare professional, policymakers and digital healthcare innovators to invent the tool which leads the health equity through the digital healthcare in the healthcare.

Due to its ability to support well-informed decision-making, business intelligence (BI) has grown in popularity among executives across a range of industries. However, given the volume of data collected in health-care organizations, there is a lack of exploration concerning its implementation. Consequently, this research paper aims to investigate the key factors affecting the acceptance and use of BI in healthcare organizations.

Leveraging the theoretical lens of the “unified theory of acceptance and use of technology” (UTAUT), a study framework was proposed and integrated with three context-related factors, including “rational decision-making culture” (RDC), “perceived threat to professional autonomy” (PTA) and “medical–legal risk” (MLR). The variables in the study framework were categorized as follows: information systems (IS) perspective; organizational perspective; and user perspective. In Jordan, 434 healthcare professionals participated in a cross-sectional online survey that was used to collect data.

The findings of the “structural equation modeling” revealed that professionals’ behavioral intentions toward using BI systems were significantly affected by performance expectancy, social influence, facilitating conditions, MLR, RDC and PTA. Also, an insignificant effect of PTA on PE was found based on the results of statistical analysis. These variables explained 68% of the variance (R²) in the individuals’ intentions to use BI-based health-care systems.

To promote the acceptance and use of BI technology in health-care settings, developers, designers, service providers and decision-makers will find this study to have a number of practical implications. Additionally, it will support the development of effective strategies and BI-based health-care systems based on these study results, attracting the interest of many users.

To the best of the author’s knowledge, this is one of the first studies that integrates the UTAUT model with three contextual factors (RDC, PTA and MLR) in addition to examining the suggested framework in a developing nation (Jordan). This study is one of the few in which the users’ acceptance behavior of BI systems was investigated in a health-care setting. More specifically, to the best of the author’s knowledge, this is the first study that reveals the critical antecedents of individuals’ intention to accept BI for health-care purposes in the Jordanian context.