PSTCNN: Explainable COVID-19 diagnosis using PSO-guided self-tuning CNN

Since 2019, the coronavirus disease-19 (COVID-19) has been spreading rapidly worldwide, posing an unignorable threat to the global economy and human health. It is a disease caused by severe acute respiratory syndrome coronavirus 2, a single-stranded RNA virus of the genus Betacoronavirus. This virus is highly infectious and relies on its angiotensin-converting enzyme 2-receptor to enter cells. With the increase in the number of confirmed COVID-19 diagnoses, the difficulty of diagnosis due to the lack of global healthcare resources becomes increasingly apparent. Deep learning-based computer-aided diagnosis models with high generalisability can effectively alleviate this pressure. Hyperparameter tuning is essential in training such models and significantly impacts their final performance and training speed. However, traditional hyperparameter tuning methods are usually time-consuming and unstable. To solve this issue, we introduce Particle Swarm Optimisation to build a PSO-guided Self-Tuning Convolution Neural Network (PSTCNN), allowing the model to tune hyperparameters automatically. Therefore, the proposed approach can reduce human involvement. Also, the optimisation algorithm can select the combination of hyperparameters in a targeted manner, thus stably achieving a solution closer to the global optimum. Experimentally, the PSTCNN can obtain quite excellent results, with a sensitivity of 93.65% ± 1.86%, a specificity of 94.32% ± 2.07%, a precision of 94.30% ± 2.04%, an accuracy of 93.99% ± 1.78%, an F1-score of 93.97% ± 1.78%, Matthews Correlation Coefficient of 87.99% ± 3.56%, and Fowlkes-Mallows Index of 93.97% ± 1.78%. Our experiments demonstrate that compared to traditional methods, hyperparameter tuning of the model using an optimisation algorithm is faster and more effective. © 2023 Centro Regional de Invest. Cientif. y Tecn.. All rights reserved.

Design of a machine learning to classify health beverages preferences for elderly people: an empirical study during COVID-19 in Thailand

This research designed a decision support system based upon a machine learning (DSS-ML) model for classifying health beverage preferences for elderly people. A neural network was designed involving training using particle swarm optimisation (PSO) in comparison with two ML models: logistic regression (LR) and a neural network (NN). The DSS-ML model was able to classify accurately and autonomously the preference complexities associated with the health beverage preferences for elderly people in accordance with the WHO's recommendation. In terms of contribution, the results demonstrated that NN training with PSO resulted in a higher ability to classify the preferences for health beverages than for the two ML models. Furthermore, NN training with PSO achieved faster convergence than NN. The benefits of this research can be separated into two parts. First, manufacturers can introduce beverages that satisfy elderly people's preferences. Second, elderly people can be made aware of appropriate health beverages. Copyright © 2022 Inderscience Enterprises Ltd.

PSTCNN: Explainable COVID-19 diagnosis using PSO-guided self-tuning CNN

Since 2019, the coronavirus disease-19 (COVID-19) has been spreading rapidly worldwide, posing an unignorable threat to the global economy and human health. It is a disease caused by severe acute respiratory syndrome coronavirus 2, a single-stranded RNA virus of the genus Betacoronavirus. This virus is highly infectious and relies on its angiotensin-converting enzyme 2-receptor to enter cells. With the increase in the number of confirmed COVID-19 diagnoses, the difficulty of diagnosis due to the lack of global healthcare resources becomes increasingly apparent. Deep learning-based computer-aided diagnosis models with high generalisability can effectively alleviate this pressure. Hyperparameter tuning is essential in training such models and significantly impacts their final performance and training speed. However, traditional hyperparameter tuning methods are usually time-consuming and unstable. To solve this issue, we introduce Particle Swarm Optimisation to build a PSO-guided Self-Tuning Convolution Neural Network (PSTCNN), allowing the model to tune hyperparameters automatically. Therefore, the proposed approach can reduce human involvement. Also, the optimisation algorithm can select the combination of hyperparameters in a targeted manner, thus stably achieving a solution closer to the global optimum. Experimentally, the PSTCNN can obtain quite excellent results, with a sensitivity of 93.65% ± 1.86%, a specificity of 94.32% ± 2.07%, a precision of 94.30% ± 2.04%, an accuracy of 93.99% ± 1.78%, an F1-score of 93.97% ± 1.78%, Matthews Correlation Coefficient of 87.99% ± 3.56%, and Fowlkes-Mallows Index of 93.97% ± 1.78%. Our experiments demonstrate that compared to traditional methods, hyperparameter tuning of the model using an optimisation algorithm is faster and more effective. [ FROM AUTHOR]

An integrated discrete event simulation and particle swarm optimisation model for optimising efficiency of cancer diagnosis pathways

The National Health Service (NHS) constitution sets out minimum standards for rights of access of patients to NHS services. The ‘Faster Diagnosis Standard’ (FDS) states that 75% of patients should be told whether they have a diagnosis of cancer or not within 28 days of an urgent GP referral. Timely diagnosis and treatment lead to improved outcomes for cancer patients, however, compliance with these standards has recently been challenged, particularly in the context of operational pressures and resource constraints relating to the COVID-19 pandemic. In order to minimise diagnostic delays, the National Physical Laboratory in collaboration with the Royal Free London (RFL) NHS Foundation Trust address this problem by treating it as a formal resource optimisation, aiming to minimise the number of patients who breach the FDS. We use discrete event simulation and particle swarm optimisation to identify areas for improving the efficiency of cancer diagnosis at the RFL. We highlight capacity-demand mismatches in the current cancer diagnosis pathways at the RFL, including imaging and endoscopy investigations. This is due to the volume of patients requiring these investigations to meet the 28-day FDS target. We find that increasing resources in one area alone does not fully solve the problem. By looking at the system as a whole we identify areas for improvement which will have system-wide impact even though individually they do not necessarily seem significant. The outcomes and impact of this project have the potential to make a valuable impact on shaping future hospital activity.

An integrated model of aircraft routing and crew rostering problems to develop fair schedule for the crew under COVID-19 condition

This paper focuses on aircraft routing and crew rostering problems simultaneously considering the risk of COVID-19 infection. As airports are among high-risk places in COVID-19 pandemic, the crew prefer to spend less sit time in airports and come back to their home base at the end of each duty day. In this research, an integrated model is developed to assign crew and aircraft to flights in order to achieve a fair schedule for the crew. The objective function is minimisation of the difference between crew sit times. Moreover in this model, a framework including flight hours, number of days and number of take-offs is considered for maintenance requirements. Particle swarm optimisation (PSO) is used as the solution approach. To validate the solution approach, 20 test problems were solved using GAMS and PSO. The results show that PSO improved CPU time significantly (98.279% in average) in turn of 1.902% gap with GAMS in optimum solution.

Comparison of health-oriented cross-regional allocation strategies for the COVID-19 vaccine: a mathematical modelling study.

BACKGROUND: Controlling the epidemic spread and establishing the immune barrier in a short time through accurate vaccine demand prediction and optimised vaccine allocation strategy are still urgent problems to be solved under the condition of frequent virus mutations. METHODS: A cross-regional Susceptible-Exposed-Infected-Removed dynamic model was used for scenario simulation to systematically elaborate and compare the effects of different cross-regional vaccine allocation strategies on the future development of the epidemic in regions with different population sizes, prevention and control capabilities, and initial risk levels. Furthermore, the trajectory of the cross-regional vaccine allocation strategy, calculated using a particle swarm optimisation algorithm, was compared with the trajectories of other strategies. RESULTS: By visualising the final effect of the particle swarm optimisation vaccine allocation strategy, this study revealed the important role of prevention and control (including the level of social distancing control, the speed of tracking and isolating exposed and infected individuals, and the initial frequency of mask-wearing) in determining the allocation of vaccine resources. Most importantly, it supported the idea of prioritising control in regions with a large population and low initial risk level, which broke the general view that high initial risk needs to be given priority and proposed that outbreak risk should be firstly considered instead. CONCLUSIONS: This is the first study to use a particle swarm optimisation algorithm to study the cross-regional allocation of COVID-19 vaccines. These data provide a theoretical basis for countries and regions to develop more targeted and sustainable vaccination strategies.KEY MESSAGEThe innovative combination of particle swarm optimisation and cross-regional SEIR model to simulate the pandemic trajectory and predict the vaccine demand helped to speed up and stabilise the construction of the immune barrier, especially faced with new virus mutations.We proposed that priority should be given to regions where it is possible to prevent more infections rather than regions where it is at high initial risk, thus regional outbreak risk should be considered when making vaccine allocation decisions.An optimal health-oriented strategy for vaccine allocation in the COVID-19 pandemic is determined considering both pharmaceutical and non-pharmaceutical policy interventions, including speed of isolation, degree of social distancing control, and frequency of mask-wearing.

Cloud- and Fog-Integrated Smart Grid Model for Efficient Resource Utilisation.

The smart grid (SG) is a contemporary electrical network that enhances the network's performance, reliability, stability, and energy efficiency. The integration of cloud and fog computing with SG can increase its efficiency. The combination of SG with cloud computing enhances resource allocation. To minimise the burden on the Cloud and optimise resource allocation, the concept of fog computing integration with cloud computing is presented. Fog has three essential functionalities: location awareness, low latency, and mobility. We offer a cloud and fog-based architecture for information management in this study. By allocating virtual machines using a load-balancing mechanism, fog computing makes the system more efficient (VMs). We proposed a novel approach based on binary particle swarm optimisation with inertia weight adjusted using simulated annealing. The technique is named BPSOSA. Inertia weight is an important factor in BPSOSA which adjusts the size of the search space for finding the optimal solution. The BPSOSA technique is compared against the round robin, odds algorithm, and ant colony optimisation. In terms of response time, BPSOSA outperforms round robin, odds algorithm, and ant colony optimisation by 53.99 ms, 82.08 ms, and 81.58 ms, respectively. In terms of processing time, BPSOSA outperforms round robin, odds algorithm, and ant colony optimisation by 52.94 ms, 81.20 ms, and 80.56 ms, respectively. Compared to BPSOSA, ant colony optimisation has slightly better cost efficiency, however, the difference is insignificant.

Review of fractional epidemic models.

The global impact of corona virus (COVID-19) has been profound, and the public health threat it represents is the most serious seen in a respiratory virus since the 1918 influenza A(H1N1) pandemic. In this paper, we have focused on reviewing the results of epidemiological modelling especially the fractional epidemic model and summarized different types of fractional epidemic models including fractional Susceptible-Infective-Recovered (SIR), Susceptible-Exposed-Infective-Recovered (SEIR), Susceptible-Exposed-Infective-Asymptomatic-Recovered (SEIAR) models and so on. Furthermore, we propose a general fractional SEIAR model in the case of single-term and multi-term fractional differential equations. A feasible and reliable parameter estimation method based on modified hybrid Nelder-Mead simplex search and particle swarm optimisation is also presented to fit the real data using fractional SEIAR model. The effective methods to solve the fractional epidemic models we introduced construct a simple and effective analytical technique that can be easily extended and applied to other fractional models, and can help guide the concerned bodies in preventing or controlling, even predicting the infectious disease outbreaks.

PSO-Blockchain-based image steganography: towards a new method to secure updating and sharing COVID-19 data in decentralised hospitals intelligence architecture.

Secure updating and sharing for large amounts of healthcare information (such as medical data on coronavirus disease 2019 [COVID-19]) in efficient and secure transmission are important but challenging in communication channels amongst hospitals. In particular, in addressing the above challenges, two issues are faced, namely, those related to confidentiality and integrity of their health data and to network failure that may cause concerns about data availability. To the authors' knowledge, no study provides secure updating and sharing solution for large amounts of healthcare information in communication channels amongst hospitals. Therefore, this study proposes and discusses a novel steganography-based blockchain method in the spatial domain as a solution. The novelty of the proposed method is the removal and addition of new particles in the particle swarm optimisation (PSO) algorithm. In addition, hash function can hide secret medical COVID-19 data in hospital databases whilst providing confidentiality with high embedding capacity and high image quality. Moreover, stego images with hash data and blockchain technology are used in updating and sharing medical COVID-19 data between hospitals in the network to improve the level of confidentiality and protect the integrity of medical COVID-19 data in grey-scale images, achieve data availability if any connection failure occurs in a single point of the network and eliminate the central point (third party) in the network during transmission. The proposed method is discussed in three stages. Firstly, the pre-hiding stage estimates the embedding capacity of each host image. Secondly, the secret COVID-19 data hiding stage uses PSO algorithm and hash function. Thirdly, the transmission stage transfers the stego images based on blockchain technology and updates all nodes (hospitals) in the network. As proof of concept for the case study, the authors adopted the latest COVID-19 research published in the Computer Methods and Programs in Biomedicine journal, which presents a rescue framework within hospitals for the storage and transfusion of the best convalescent plasma to the most critical patients with COVID-19 on the basis of biological requirements. The validation and evaluation of the proposed method are discussed.