Applied Artificial Intelligence: Medicine, Biology, Chemistry, Financial, Games, Engineering

Preface
Contents
Advances in the Use of Artificial Intelligence and Sensor Technologies for Managing Industrial Workplace Safety
	1 Introduciton
		1.1 Workplace Safety Management in SMEs
		1.2 The Importance of Timely and Objective Identification of UC/UA
	2 Misuse of PPE as the Use Case of Unsafe Acts
	3 The Use of AI for Assessing the Safety of Pushing and Pulling Activities
		3.1 Workplace Musculoskeletal Disorders and Injuries
		3.2 Computer Vision, Deep Learning and Workplace Safety
		3.3 The Use of Sensors for Analyzing Workplace Safety
		3.4 The Use of 3D Pose Estimation and Human Body Models
	4 Assessment of the Human–Robot Collaborative Polishing Task by Using EMG Sensors and 3D Pose Estimation
	5 The Use of EEG for Workplace Safety Assessment
		5.1 Development of Modular and Adaptive Laboratory Set-Up for Neuroergonomic and Human–Robot Interaction Research
	6 Influence of Operators’ Psychological and Physiological Characteristics on Workplace Safety
	7 Conclusions
	References
Implementation of Deep Learning to Prevent Peak-Driven Power Outages Within Manufacturing Systems
	1 Introduction
		1.1 The Role of IoT, Big Data, and AI in Managing Manufacturing Systems
	2 Research Approach, Aim, and Structure
		2.1 Manufacturing System Overview
		2.2 IoT Device Overview
		2.3 AI Model Development
	3 Obtained Results
		3.1 Conceptual Deployment Approach
	4 Conclusions
	References
Reproductive Autonomy Conformity Assessment of Purposed AI System
	1 Introduction
	2 A State-of-Art and Legal Landscape for AI Application
	3 A Reproductive Autonomy Conformity Assessment
		3.1 Implantation Data and Reproductive Autonomy
		3.2 AI and Informed Consent
		3.3 Reproductive Autonomy and Human Intervention
	4 Conclusion
	5 Summary
	References
Baselines for Automatic Medical Image Reporting
	1 Introduction
	2 Materials
		2.1 Pre-processing
	3 Experimentations
		3.1 Normal vs Non-normal
		3.2 Fine-Tuning CNNs on IU-CHEST
		3.3 Decoders
	4 Discussion
	5 Conclusions
	References
HR Analytics: Serbian Perspective
	1 Introduction
	2 HR Analytics
		2.1 HR X.0
	3 HR Analytics Framework
	4 Step Forward to Industry 5.0 and HR 5.0
	5 CS: Serbia
	6 Conclusions and Recommendations
		6.1 Recommendations for Action
	References
Ontology-Based Analysis of Job Offers for Medical Practitioners in Poland
	1 Introduction
	2 Research Methodology
		2.1 Job Offers Retrieving
		2.2 Semantic Annotation
		2.3 Bipartite Graph Models
	3 Analysis of Job Offers for Medical Practitioners in Poland
		3.1 General Description of the Research Process
		3.2 Analysis of the Specialization-Locations Bipartite Model
		3.3 Analysis of the Specialization-Institutions Bipartite Model
	4 Conclusions
	References
Synergizing Four Different Computing Paradigms for Machine Learning and Big Data Analytics
	1 Introduction
	2 Comparison of Four Computing Paradigms
	3 Possible Architecture of a Supercomputer on a Chip
	4 Elaboration
	5 Conclusion
	References
Pose Estimation and Joint Angle Detection Using Mediapipe Machine Learning Solution
	1 Introduction
	2 Methods
		2.1 Input Data
		2.2 MediaPipe Pose
		2.3 Angle Detection
		2.4 Design Scaling
		2.5 Hardware and Software Requirements
	3 Results
	4 Conclusions
	References
Application of AI in Histopathological Image Analysis
	1 Introduction
		1.1 Related Work
	2 Materials and Methods
		2.1 Dataset Description
		2.2 Convolutional Neural Network Architectures
	3 Results and Discussion
	4 Conclusions
	References
The Projects Evaluation and Selection by Using MCDM and Intuitionistic Fuzzy Sets
	1 Introduction
	2 Problem Statement
		2.1 Modelling of Uncertainties
	3 Methodology
		3.1 The Proposed Algorithm
	4 An Illustrative Example
	5 Conclusions
	Appendix A: Input data
	Appendix B: Preliminaries
	References
Application of MCDM DIBR-Rough Mabac Model for Selection of Drone for Use in Natural Disaster Caused by Flood
	1 Introduction
	2 Description of Model and Used Methods
		2.1 DIBR Method
		2.2 Rough Numbers
		2.3 Rough MABAC Method
	3 Application of MCDM Model and Results
	4 Sensitivity Analysis
	5 Conclusions
	References
Improving the Low Accuracy of Traditional Earthquake Loss Assessment Systems
	1 Introduction
	2 The Traditional Approach to Rapid Earthquake Loss Assessment and What’s Wrong with It?
		2.1 Pre-earthquake Phase – Preparing the System Before an Earthquake
		2.2 Co-earthquake Phase - Activating the System Immediately After an Earthquake
	3 The New Approach and Framework for Rapid Earthquake Loss Assessment
		3.1 M5.4 2010 Kraljevo Earthquake Data Modeling
		3.2 Can ML Be Used for Earthquake Damage Prediction and How?
		3.3 Do We Even Need Earthquake Data to Predict Earthquake Damage?
		3.4 Improving Monetizing Loss - Repair Cost Matrix and the “Soft Rule”
		3.5 Representative Sampling
		3.6 The Proposed RELA (Rapid Earthquake Loss Assessment) Framework
		3.7 Achieved Accuracy Using the RELA Framework
	4 Beyond Loss Assessment – The Recovery Process
	5 Conclusions
	References
SecondOponionNet: A Novel Neural Network Architecture to Detect Coronary Atherosclerosis in Coronary CT Angiography
	1 Introduction
	2 Materials and Methods
		2.1 Dataset
		2.2 Methodology
	3 Results and Discussion
	4 Conclusion
	References
Ontology-Based Exploratory Text Analysis as a Tool for Identification of Research Trends in Polish Universities of Economics
	1 Introduction
	2 The Theoretical Background
		2.1 Challenges of Contemporary Publication Analysis
		2.2 Characteristics of Research Publication
	3 Methodology
		3.1 Research Methodology
		3.2 Ontology-Based Analysis of Publication Abstracts
		3.3 Graph-Based Model of Publication Activity
	4 The Analysis of the Publication Activity of Researchers from Polish Universities of Economics
		4.1 Publication Activity of Researchers from Polish Universities of Economics
		4.2 Analysis of JEL Category Importance
		4.3 Analysis of Connections Between JEL Categories and Universities
	5 Conclusions
	References
Improved Three-Dimensional Reconstruction of Patient-Specific Carotid Bifurcation Using Deep Learning Based Segmentation of Ultrasound Images
	1 Introduction
	2 Materials and Methods
		2.1 Deep Learning Segmentation
		2.2 Improved 3D Reconstruction
		2.3 Fluid Flow Simulations
	3 Results
	4 Conclusions
	References
Seat-to-Head Transfer Functions Prediction Using Artificial Neural Networks
	1 Introduction
		1.1 Literature According to the Type of Vibration Excitation
		1.2 Literature According to the Place of Vibration Examination
		1.3 Literature According to the Direction of Action of the Vibration Excitation
		1.4 Literature According to Human Body Position
	2 Seat-to-Head Transmissibility Function
		2.1 Influencing Factors on STHT Function
	3 Neural Networks
		3.1 A Multi-layer Perceptron
		3.2 Recurrent Neural Networks
		3.3 LSTM
	4 Methods
	5 Results
		5.1 Experimental Measurements
		5.2 Development of ANN Prediction Models
	6 Conclusions
	References
A Review of the Application of Artificial Intelligence in Medicine: From Data to Personalised Models
	1 Introduction
	2 Application of Artificial Intelligence in Medicine
		2.1 Stratification of Patients with Carotid Artery Disease
		2.2 Assessment of Patient Condition with Familiar Cardiomyopathy
		2.3 Personalized COVID-19 Model
		2.4 Epidemiological COVID-19 Model
		2.5 Integration of Different Models into Multiscale Platform
	3 AI in Medicine – Current Limitations and Future Trends
	4 Conclusions
	References
Digital Platform as the Communication Channel for Challenges in Artificial Intelligence
	1 Introduction
	2 Platform Concept
	3 Backend, Frontend and Discussion
		3.1 Backend
		3.2 Frontend
		3.3 Public
		3.4 Company, Investor or Research Group
		3.5 Solvers
		3.6 Reviewer
		3.7 Administrator
		3.8 Super Administrator
	4 Conclusion
	References
Mathematical Modeling of COVID-19 Spread Using Genetic Programming Algorithm
	1 Introduction
	2 Materials and Methods
	3 Results and Discussion
	4 Conclusions
	References
Liver Tracking for Intraoperative Augmented Reality Navigation System
	1 Introduction
		1.1 Virtual and Augmented Reality in Medicine
		1.2 Augmented Reality for Liver Resection
	2 Materials and Methods
		2.1 Manual Registration
		2.2 Semi-automatic Registration
		2.3 Automatic Registration and Liver Tracking
	3 Results
	4 Conclusion
	References
Intelligent Drug Delivery Systems
	1 Introduction
	2 Principles of Controlled Drug Delivery
		2.1 Controlled Delivery
		2.2 Delayed or Initiated Delivery
		2.3 Targeted Delivery
		2.4 Endogenous and Exogenous Stimuli for Drug Release
	3 Modern Drug Delivery Systems
		3.1 Tablets
		3.2 Patches
		3.3 Liposomes
		3.4 Nanoparticles
	4 Materials for Controlled Drug Delivery
		4.1 Poly (Lactide - Co - Glycolide)
		4.2 Gelatin
		4.3 Collagen
	5 Fabrication of Drug Delivery Systems
		5.1 Conventional Tablet Manufacturing
		5.2 Tablet Manufacturing Using 3D Printing
		5.3 Electrospinning
		5.4 Production of Nanoparticles
		5.5 Production of Liposomes
	6 Selection of Medicine
	7 Smart Devices for Controlled Drug Delivery
		7.1 Microchips
		7.2 Ophthalmological Devices
		7.3 Transdermal Devices
		7.4 Medicine Delivery Pumps
	8 Mathematical Models and Artificial Intelligence
	9 Conclusion
	References
Cost Effectiveness Analysis of Real and in Silico Clinical Trials for Stent Deployment Based on Decision Tree
	1 Introduction
	2 Stent Market
		2.1 Coronary Stents
		2.2 Peripheral Stents Market
		2.3 InSilc Platform
	3 Mechanical Modelling Module
		3.1 Deployment Module
	4 Cost Effectiveness for the InSilc Platform
		4.1 Mechanical Modelling Module Pricing Strategy
		4.2 Scenario 1 – Pre-clinical Testing Assessment
		4.3 Scenario 2 - Design New Stents
		4.4 Scenario 3 - Compare Existing Stents
		4.5 Scenario 4 - Compare Anatomy Configurations and Patient Conditions
		4.6 Scenario 5 - Compare Different Revascularization Procedures
	5 Surface Reconstruction Based on Stent Deployment Simulation and 3D Imaging Data
	6 Cost Decision Tree Calculation
	7 Conclusion
	References
Author Index