There has been an increasing interest in translating artificial intelligence (AI) research into clinically-validated applications to improve the performance, capacity, and efficacy of healthcare services. Despite substantial research worldwide, very few AI-based applications have successfully made it to clinics. Key barriers to the widespread adoption of clinically validated AI applications include non-standardized medical records, limited availability of curated datasets, and stringent legal/ethical requirements to preserve patients' privacy. Therefore, there is a pressing need to improvise new data-sharing methods in the age of AI that preserve patient privacy while developing AI-based healthcare applications. In the literature, significant attention has been devoted to developing privacy-preserving techniques and overcoming the issues hampering AI adoption in an actual clinical environment. To this end, this study summarizes the state-of-the-art approaches for preserving privacy in AI-based healthcare applications. Prominent privacy-preserving techniques such as Federated Learning and Hybrid Techniques are elaborated along with potential privacy attacks, security challenges, and future directions.

Recent years have witnessed widespread adoption of machine learning (ML)/deep learning (DL) techniques due to their superior performance for a variety of healthcare applications ranging from the prediction of cardiac arrest from one-dimensional heart signals to computer-aided diagnosis (CADx) using multi-dimensional medical images. Notwithstanding the impressive performance of ML/DL, there are still lingering doubts regarding the robustness of ML/DL in healthcare settings (which is traditionally considered quite challenging due to the myriad security and privacy issues involved), especially in light of recent results that have shown that ML/DL are vulnerable to adversarial attacks. In this paper, we present an overview of various application areas in healthcare that leverage such techniques from security and privacy point of view and present associated challenges. In addition, we present potential methods to ensure secure and privacy-preserving ML for healthcare applications. Finally, we provide insight into the current research challenges and promising directions for future research.

Purpose: In a circular economy, the goal is to keep materials values in the economy for as long as possible. For the construction industry to support the goal of the circular economy, there is the need for materials reuse. However, there is little or no information about the amount and quality of reusable materials obtainable when buildings are deconstructed. The purpose of this paper, therefore, is to develop a reusability analytics tool for assessing end-of-life status of building materials. Design/methodology/approach: A review of the extant literature was carried out to identify the best approach to modelling end-of-life reusability assessment tool. The reliability analysis principle and materials properties were used to develop the predictive mathematical model for assessing building materials performance. The model was tested using the case study of a building design and materials take-off quantities as specified in the bill of quantity of the building design. Findings: The results of analytics show that the quality of the building materials varies with the building component. For example, from the case study, at the 80th year of the building, the qualities of the obtainable concrete from the building are 0.9865, 0.9835, 0.9728 and 0.9799, respectively, from the foundation, first floor, frame and stair components of the building. Originality/value: As a contribution to the concept of circular economy in the built environment, the tool provides a foundation for estimating the quality of obtainable building materials at the end-of-life based on the life expectancy of the building materials.

Purpose: The purpose of this paper is to highlight the use of the big data technologies for health and safety risks analytics in the power infrastructure domain with large data sets of health and safety risks, which are usually sparse and noisy. Design/methodology/approach: The study focuses on using the big data frameworks for designing a robust architecture for handling and analysing (exploratory and predictive analytics) accidents in power infrastructure. The designed architecture is based on a well coherent health risk analytics lifecycle. A prototype of the architecture interfaced various technology artefacts was implemented in the Java language to predict the likelihoods of health hazards occurrence. A preliminary evaluation of the proposed architecture was carried out with a subset of an objective data, obtained from a leading UK power infrastructure company offering a broad range of power infrastructure services. Findings: The proposed architecture was able to identify relevant variables and improve preliminary prediction accuracies and explanatory capacities. It has also enabled conclusions to be drawn regarding the causes of health risks. The results represent a significant improvement in terms of managing information on construction accidents, particularly in power infrastructure domain. Originality/value: This study carries out a comprehensive literature review to advance the health and safety risk management in construction. It also highlights the inability of the conventional technologies in handling unstructured and incomplete data set for real-time analytics processing. The study proposes a technique in big data technology for finding complex patterns and establishing the statistical cohesion of hidden patterns for optimal future decision making.

Using 693 000 datacells from 33 000 sample construction firms that operated or failed between 2008 and 2017, failure prediction models were developed using artificial neural network (ANN), support vector machine, multiple discriminant analysis (MDA), and logistic regression (LR). The accuracy of the models on test data surprisingly showed ANN to have only a slightly better accuracy than LR and MDA. The ANN's number of units in the hidden layer and weight decay hyperparameters were consequently tuned using the grid search. Tuning process led to tedious machine computation that was aborted after many hours without completion. The state of art big data analytics (BDA) technology was, for the first time in failure prediction, consequently employed and the tuning was completed in some seconds. Mean accuracy from cross validation was used for selection of the model with best parameter values, which were used to develop a new ANN model that outperformed all previously developed models on test data. Subsequent use of selected variables to develop new models led to reduced tuning computational cost, but not improved performance. Since the real-life effect of a misclassification cost is greater than the tedious computation cost, it was concluded that BDA is the best compromise.

Construction industry insolvency studies have failed to stem the industry’s high insolvency tide because many focus on big civil engineering firms (CEF) when over 90% of firms in the industry are small or micro (S&M). This study thus set out to uncover insolvency criteria of S&M CEFs and the underlying factors using mixed methods. Using convenience sampling, the storytelling method was used to execute interviews of 16 respondents from insolvent firms. Narrative and thematic analysis were used to extract 17 criteria under 2 groups. Criteria were used to formulate a questionnaire, of which 81 completed copies were received and analyzed using Cronbach’s alpha coefficient and relevance index score for reliability and ranking, respectively. The five most relevant criteria were economic recession, immigration, too many new firms springing up, collecting receivables, and burden of sustainable construction. The four underlying factors established through factor analysis were market forces, competence-based management, operations efficiency and other management issues, and information management. The factors were in line with Mintzberg’s and Porter’s strategy theories. The results demonstrate that insolvency factors affecting big and small CEF can be quite different and, sometimes, even opposite. This research will provide a unique resource on the factors that should make potential owners of S&M CEF cautious. The criteria are potential variables for insolvency prediction models for S&M CEFs.