Explainable Machine-Learning Models for COVID-19 Prognosis Prediction Using Clinical, Laboratory and Radiomic Features DOI Creative Commons
Francesco Prinzi, Carmelo Militello, Nicola Scichilone

et al.

IEEE Access, Journal Year: 2023, Volume and Issue: 11, P. 121492 - 121510

Published: Jan. 1, 2023

The SARS-CoV-2 virus pandemic had devastating effects on various aspects of life: clinical cases, ranging from mild to severe, can lead lung failure and death. Due the high incidence, data-driven models support physicians in patient management. explainability interpretability machine-learning are mandatory scenarios. In this work, clinical, laboratory radiomic features were used train for COVID-19 prognosis prediction. Using Explainable AI algorithms, a multi-level explainable method was proposed taking into account developer involved stakeholder (physician, patient) perspectives. A total 1023 extracted 1589 Chest X-Ray images (CXR), combined with 38 clinical/laboratory features. After pre-processing selection phases, 40 CXR 23 Support Vector Machine Random Forest classifiers exploring three feature strategies. combination both radiomic, enabled higher performance resulting models. intelligibility allowed us validate models' findings. According medical literature, LDH, PaO2 CRP most predictive Instead, ZoneEntropy HighGrayLevelZoneEmphasis - indicative heterogeneity/uniformity texture discriminating Our best model, exploiting classifier signature composed features, achieved AUC=0.819, accuracy=0.733, specificity=0.705, sensitivity=0.761 test set. including explainability, allows make strong assumptions, confirmed by literature insights.

Language: Английский

Applications of machine learning in thermochemical conversion of biomass-A review DOI
Muzammil Khan, Salman Raza Naqvi, Zahid Ullah

et al.

Fuel, Journal Year: 2022, Volume and Issue: 332, P. 126055 - 126055

Published: Sept. 24, 2022

Language: Английский

Citations

141

The applications of MCDM methods in COVID-19 pandemic: A state of the art review DOI
Alireza Sotoudeh-Anvari

Applied Soft Computing, Journal Year: 2022, Volume and Issue: 126, P. 109238 - 109238

Published: June 30, 2022

Language: Английский

Citations

136

Worker and workplace Artificial Intelligence (AI) coexistence: Emerging themes and research agenda DOI Creative Commons
Araz Zirar, Syed Imran Ali, Nazrul Islam

et al.

Technovation, Journal Year: 2023, Volume and Issue: 124, P. 102747 - 102747

Published: March 15, 2023

Workplace Artificial Intelligence (AI) helps organisations increase operational efficiency, enable faster-informed decisions, and innovate products services. While there is a plethora of information about how AI may provide value to workplaces, research on workers can coexist in workplaces evolving. It critical explore emerging themes agendas understand the trajectory scholarly this area. This study's overarching question will with workplaces. A search protocol was employed find relevant articles Scopus, ProQuest, Web Science databases based appropriate specific keywords article inclusion exclusion criteria. We identified four themes: (1) Workers' distrust workplace stems from perceiving it as job threat, (2) entices worker-AI interactions by offering augment worker abilities, (3) coexistence require workers' technical, human, conceptual skills, (4) Workers need ongoing reskilling upskilling contribute symbiotic relationship AI. then developed propositions questions for future research. review makes contributions: argues that an existential argument better explains AI, gathers required skills groups them into suggests technical benefit but cannot outweigh human offers 20 evidence-informed guide inquiries.

Language: Английский

Citations

133

The Role of Artificial Intelligence and Machine Learning Amid the COVID-19 Pandemic: What Lessons Are We Learning on 4IR and the Sustainable Development Goals DOI Open Access
David Mhlanga

International Journal of Environmental Research and Public Health, Journal Year: 2022, Volume and Issue: 19(3), P. 1879 - 1879

Published: Feb. 8, 2022

The COVID-19 pandemic came with disruptions in every aspect of human existence, all the sectors economies world affected greatly. In health sector, halted and reversed progress subsequently shortened life expectancy, especially developing underdeveloped nations. On other hand, machine learning artificial intelligence contributed a great deal to handling globally. Therefore, current study aimed assess role played by addressing dangers posed pandemic, as well extrapolate lessons on fourth industrial revolution sustainable development goals. Using qualitative content analysis, results indicated that an important response challenges pandemic. Artificial intelligence, learning, various digital communication tools through telehealth performed meaningful roles scaling customer communications, provided platform for understanding how spreads, sped up research treatment COVID-19, among notable achievements. we draw from this is that, despite rise number unintended consequences technology revolution, motivates us conclude governments must build trust these technologies, address problems going forward, ensure goals related good wellbeing are achieved.

Language: Английский

Citations

94

Artificial neural networks for the prediction of biochar yield: A comparative study of metaheuristic algorithms DOI Creative Commons
Muzammil Khan, Zahid Ullah, Ondřej Mašek

et al.

Bioresource Technology, Journal Year: 2022, Volume and Issue: 355, P. 127215 - 127215

Published: April 23, 2022

In this study, an integrated framework of artificial neural networks (ANNs) and metaheuristic algorithms have been developed for the prediction biochar yield using biomass characteristics pyrolysis process conditions. Comparative analysis six different was performed to optimize ANN architecture select important features. The results suggested that model coupled with Rao-2 algorithm outperformed (R2 ∼ 0.93, RMSE 1.74%) all other models. Furthermore, detailed information behind models acquired, identifying most influencing factors as follows: temperature (56%), residence time (23%), heating rate (8%). partial dependence plot revealed how each factor affected target variable. Finally, easy-to-use software tool predicting built ANN-Rao-2 model. This study demonstrates huge potential machine learning presents in predictive modelling complex processes, reduces time-consuming expensive experimental work estimating yield.

Language: Английский

Citations

86

Application of Artificial Intelligence in Combating High Antimicrobial Resistance Rates DOI Creative Commons
Ali A. Rabaan, Saad Alhumaid, Abbas Al Mutair

et al.

Antibiotics, Journal Year: 2022, Volume and Issue: 11(6), P. 784 - 784

Published: June 8, 2022

Artificial intelligence (AI) is a branch of science and engineering that focuses on the computational understanding intelligent behavior. Many human professions, including clinical diagnosis prognosis, are greatly useful from AI. Antimicrobial resistance (AMR) among most critical challenges facing Pakistan rest world. The rising incidence AMR has become significant issue, authorities must take measures to combat overuse incorrect use antibiotics in order rates. widespread practice not only resulted drug but also increased threat super-resistant bacteria emergence. As rises, clinicians find it more difficult treat many bacterial infections timely manner, therapy becomes prohibitively costly for patients. To rise rates, implement an institutional antibiotic stewardship program monitors correct use, controls antibiotics, generates antibiograms. Furthermore, these types tools may aid treatment patients event medical emergency which physician unable wait culture results. AI's applications healthcare might be unlimited, reducing time takes discover new antimicrobial drugs, improving diagnostic accuracy, lowering expenses at same time. majority suggested AI solutions meant supplement rather than replace doctor's prescription or opinion, serve as valuable tool making their work easier. When comes infectious diseases, potential game-changer battle against resistance. Finally, when selecting infections, data local programs ensuring treated quickly effectively. organizations such World Health Organization (WHO) have underlined necessity appropriate treating shortest feasible minimize spread resistant invasive strains.

Language: Английский

Citations

84

Discovering topics and trends in the field of Artificial Intelligence: Using LDA topic modeling DOI
Dejian Yu, Bo Xiang

Expert Systems with Applications, Journal Year: 2023, Volume and Issue: 225, P. 120114 - 120114

Published: April 13, 2023

Language: Английский

Citations

67

Surveillance and response strategies for zoonotic diseases: a comprehensive review DOI Creative Commons
Manjeet Sharan, Deepthi Vijay, Jay Prakash Yadav

et al.

Science in One Health, Journal Year: 2023, Volume and Issue: 2, P. 100050 - 100050

Published: Jan. 1, 2023

Out of all emerging infectious diseases, approximately 75% are zoonotic origin, with their source often traced back to animals. The emergence zoonoses is driven by a complex interplay between anthropogenic, genetic, ecological, socioeconomic, and climatic factors. This intricate web influences poses significant challenges for the prediction prevention outbreaks. Effective coordination collaboration among animal, human, environmental health sectors essential proactively addressing major diseases. Despite advancements in surveillance diagnostic practices, continues be pressing global concern. Therefore, prioritizing disease paramount importance as part comprehensive containment strategy. Furthermore, evaluating existing systems provides insights into faced, which can mitigated through implementation 'One Health' principles involving relevant stakeholders. To initiate multisectoral partnerships, it crucial identify priorities core themes equitable inputs from various sectors. Strengthening surveillance, promoting data sharing, enhancing laboratory testing capabilities, fostering joint outbreak responses both human animal will establish necessary infrastructure effectively prevent, predict, detect, respond threats, thereby reinforcing security. review assesses approaches offering an overview agencies engaged monitoring outlines components required at human-animal-environment interface designing networks. Additionally, discusses key steps executing effective One Health approach, while highlighting encountered establishing such robust system.

Language: Английский

Citations

55

Application of artificial intelligence for resilient and sustainable healthcare system: systematic literature review and future research directions DOI
Laxmi Pandit Vishwakarma, Rajesh Kumar Singh, Ruchi Mishra

et al.

International Journal of Production Research, Journal Year: 2023, Volume and Issue: unknown, P. 1 - 23

Published: March 13, 2023

Recent years have witnessed increased pressure across the global healthcare system during COVID-19 pandemic. The pandemic shattered existing operations and taught us importance of a resilient sustainable system. Digitisation, specifically adoption Artificial Intelligence (AI) has positively contributed to developing in recent past. To understand how AI contributes building system, this study based on systematic literature review 89 articles extracted from Scopus Web Science databases is conducted. organised around several key themes such as applications, benefits, challenges using technology sector. It observed that wide applications radiology, surgery, medical, research, development Based analysis, research framework proposed an extended Antecedents, Practices, Outcomes (APO) framework. This comprises applications' antecedents, practices, outcomes for Consequently, three propositions are drawn study. Furthermore, our adopted theory, context methodology (TCM) provide future directions, which can be used reference point studies.

Language: Английский

Citations

47

Lightweight deep CNN-based models for early detection of COVID-19 patients from chest X-ray images DOI Open Access
Haval I. Hussein,

Abdulhakeem O. Mohammed,

Masoud Muhammed Hassan

et al.

Expert Systems with Applications, Journal Year: 2023, Volume and Issue: 223, P. 119900 - 119900

Published: March 18, 2023

Language: Английский

Citations

45