Applied Biomedical Engineering Using Artificial Intelligence and Cognitive Models

Description

Applied Biomedical Engineering Using Artificial Intelligence and Cognitive Models focuses on the relationship between three different multidisciplinary branches of engineering: Biomedical Engineering, Cognitive Science and Computer Science through Artificial Intelligence models. These models will be used to study how the nervous system and musculoskeletal system obey movement orders from the brain, as well as the mental processes of the information during cognition when injuries and neurologic diseases are present in the human body. The interaction between these three areas are studied in this book with the objective of obtaining AI models on injuries and neurologic diseases of the human body, studying diseases of the brain, spine and the nerves that connect them with the musculoskeletal system. There are more than 600 diseases of the nervous system, including brain tumors, epilepsy, Parkinson's disease, stroke, and many others. These diseases affect the human cognitive system that sends orders from the central nervous system (CNS) through the peripheral nervous systems (PNS) to do tasks using the musculoskeletal system. These actions can be detected by many Bioinstruments (Biomedical Instruments) and cognitive device data, allowing us to apply AI using Machine Learning-Deep Learning-Cognitive Computing models through algorithms to analyze, detect, classify, and forecast the process of various illnesses, diseases, and injuries of the human body. Applied Biomedical Engineering Using Artificial Intelligence and Cognitive Models provides readers with the study of injuries, illness, and neurological diseases of the human body through Artificial Intelligence using Machine Learning (ML), Deep Learning (DL) and Cognitive Computing (CC) models based on algorithms developed with MATLAB® and IBM Watson®. Front Cover Applied Biomedical Engineering Using Artificial Intelligence and Cognitive Models Copyright Page Dedication Contents About the author Foreword Preface Acknowledgment 1 Biomedical engineering and the evolution of artificial intelligence 1.1 Introduction 1.2 Biomedical engineering 1.2.1 Main purposes of AI in biomedical engineering 1.2.2 AI and biomedical engineering help in medical education 1.3 Artificial intelligence 1.3.1 Turing test/Turing machine 1.3.2 Basic types of AI systems based on capabilities 1.3.3 Basic types of AI systems based on functionality 1.3.4 AI technology evolution 1.3.5 AI in industries 1.4 Machine learning 1.4.1 ML seven specific steps 1.4.1.1 Step 1) Data collection 1.4.1.2 Step 2) Data preparation and exploration 1.4.1.3 Step 3) Feature engineering 1.4.1.4 Step 4) Model selection 1.4.1.5 Step 5) Model training 1.4.1.6 Step 6) Model evaluation and tuning 1.4.1.7 Step 7) Model prediction and deployment 1.5 Deep learning 1.5.1 Difference between deep learning and machine learning 1.5.2 Types of artificial neural networks 1.5.3 Feed forward neural network 1.5.4 Backpropagation neural network 1.5.5 Recurrent neural networks 1.5.6 Memory augmented neural networks 1.5.7 Modular neural networks 1.5.8 Evolutionary deep neural networks 1.6 Cognitive science 1.6.1 Cognitive computing 1.6.2 Signal recognition instruments* 1.6.3 Cognitive detection of human-like abilities* 1.6.4 Cognitive computing applying AI technologies* 1.6.5 Cognitive model obtainment* 1.6.5.1 Cognitive evaluations of human functionalities 1.6.6 Inference engine* 1.6.6.1 Knowledge storage AI database* 1.6.6.2 Action generation* 1.6.6.3 Output devices for signal generation* 1.7 Neuroscience, cognitive science, and AI models 1.7.1 The near future of neuroscience, cognitive science, and AI-ML-DL-CC References 2 Introduction to Cognitive Science, Cognitive Computing, and Human Cognitive relation to help in the solution of Artificia... 2.1 Introduction 2.2 Brain, spinal cord, and nerves 2.3 Neurons and neural pathways in cognition 2.3.1 Neurons and cognition 2.3.2 Neural pathway and cognition 2.3.3 Dopamine pathways and cognition 2.4 Cognitive science 2.5 Natural Language Processing 2.5.1 Step 1) Reading dataset for NLP 2.5.2 Step 2) Preprocessing text for NLP 2.5.3 Step 3) Data vectorization in NLP 2.5.4 Step 4) Feature engineering in NLP 2.5.5 Step 5) ML model selection for NLP 2.6 MATLAB® toolboxes solution for natural language processing 2.6.1 Natural Language Processing applications with MATLAB 2.6.2 “NLP Topic Models” with MATLAB 2.6.2.1 Research 2.1: “NLP Topic Models between a blog with a patient with neurologic disease and a researcher” 2.6.2.1.1 General objective 2.6.2.1.2 Specific objectives of this research 2.6.2.1.3 Developing a MATLAB program for this research: “NLP Topic Models” 2.6.2.1.4 Results from the MATLAB program for research: “NLP Topic Models” 2.6.2.1.5 Conclusions and recommendation 2.6.3 “NLP audio files” with MATLAB 2.6.3.1 Research 2.2: “NLP read and reproduce audio files stored and by frame in real time using MATLAB” 2.6.3.1.1 General objective 2.6.3.1.2 Specific objectives of this research 2.6.3.1.3 Developing a MATLAB program for this research: “NLP MATLAB audio files” 2.6.3.1.4 Results of the MATLAB program for “NLP read audio files” 2.6.3.1.5 Conclusions and recommendation for research 2.6.4 “NLP Text to Speech” using MATLAB 2.6.4.1 Research 2.3: “NLP Text to Speech” as action generation using MATLAB 2.6.4.1.1 General objective 2.6.4.1.2 Specific objectives of this research 2.6.4.1.3 Procedure 2.6.4.1.4 Results from “NLP Text to Speech” with MATLAB 2.6.4.1.5 Conclusions and recommendation for research: “NLP Text to Speech” with MATLAB 2.6.5 “NLP Speech to Text” with MATLAB and IBM Cloud API 2.6.5.1 Research 2.4: “NLP Text to Speech” as action generation using MATLAB and IBM Cloud API 2.6.5.1.1 General objective 2.6.5.1.2 Specific objectives of this research 2.6.5.1.3 Developing a MATLAB program for this research 2.6.5.1.4 Conclusion and recommendation for this research 2.7 Cloud service and AI 2.7.1 Cloud service providers and AI 2.8 IBM Cloud, IBM Watson, and Cognitive apps 2.8.1 IBM Cloud solution for natural language processing 2.8.2 IBM Cloud exercise to create APIs for NLP applications 2.8.2.1 General objective 2.8.2.2 Specific objectives of this example 2.8.2.3 Research 2.5: “Creating more IBM Cloud API services and testing them using command lines with cURL as an open software” 2.8.2.3.1 General objective 2.8.2.3.2 Specific objectives of this research 2.8.2.3.3 Developing IBM Cloud NLP applications with IBM APIs 2.8.2.3.4 Conclusions 2.9 The future of the relationship between cognitive science, cognitive computing, and human cognition References Further reading 3 Artificial Intelligence Models Applied to Biomedical Engineering 3.1 Introduction artificial intelligence and biomedical engineering 3.2 AI optimization in biomedical engineering 3.3 Evolutionary algorithms for AI optimization in BME 3.3.1 A typical evolutionary algorithm 3.3.2 Genetic algorithms for AI optimization in BME 3.3.2.1 Research 3.1 Genetic algorithm basic seven steps for selection by priorities 3.3.2.1.1 Problem 3.3.2.1.2 Objective 3.3.2.1.3 Procedure 3.3.3 Genetic algorithm for AI optimization in BME under MATLAB® 3.3.3.1 Research 3.2 Implementing genetic algorithm for AI optimization in BME with MATLAB 3.3.3.1.1 Problem 3.3.3.1.2 General objective 3.3.3.1.3 Specific objectives 3.3.3.1.4 Background 3.3.3.1.5 Dataset 3.3.3.1.6 Procedure 3.3.3.1.7 Results 3.3.4 General analysis and optimization of 2D and 3D data in biomedical engineering 3.3.5 MATLAB analysis and optimization of “2D” data in biomedical engineering 3.3.5.1 Research 3.3 Analysis and optimization of “2D” data of “Body measurement of nerve contractions” applying MATLAB 3.3.5.1.1 Problem to resolve 3.3.5.1.2 General objective 3.3.5.1.3 Specific objectives 3.3.5.1.4 Dataset 3.3.5.1.5 Procedure 3.3.6 MATLAB analysis and optimization of 3D data in biomedical engineering 3.3.6.1 Research 3.4 Analysis and optimization of “3D” data for “the center of mass of an upper extremity—right arm movemen... 3.3.6.1.1 Problem to resolve 3.3.6.1.2 General objective 3.3.6.1.3 Specific objectives 3.3.6.1.4 Procedure 3.4 IBM Watson Studio for artificial intelligence 3.4.1 IBM SPSS Modeler Flow 3.4.2 IBM Watson using SPSS Modeler Flow for general dataset analysis 3.4.2.1 Research 35 IBM Watson using SPSS Modeler Flow for “diabetes” analysis 3.4.2.1.1 General objective 3.4.2.1.2 Specific objectives 3.4.2.1.3 Dataset 3.4.2.1.4 Procedure 3.5 Examples of applications of evolutionary algorithms with other AI tools in biomedical engineering References 4 Machine Learning Models Applied to Biomedical Engineering 4.1 Introduction 4.2 Choosing the best ML model 4.2.1 Unsupervised learning 4.2.2 Supervised learning 4.2.3 Reinforcement learning 4.2.4 Survival models 4.2.5 Association Rules 4.3 ML clusters, classification, and regression models 4.4 Naive Bayes family models for supervised learning 4.4.1 Models: Gaussian Naive Bayes, multinomial Naive Bayes, Bernoulli Naive Bayes, Kernel Naive Bayes 4.5 k-Nearest neighbor family models for supervised learning 4.5.1 Family models: fine kNN, medium kNN, coarse kNN, cosine kNN, cubic kNN, and weighted kNN 4.6 Decision trees family models for supervised learning 4.6.1 Family models: fine decision tree, medium decision tree, and coarse decision tree 4.7 Support vector machine family members 4.7.1 Family models: linear SVM, fine Gaussian SVM, medium Gaussian SVM, coarse Gaussian SVM, quadratic SVM, and cubic SVM 4.8 Artificial neural network family models 4.8.1 Feed forward neural network family models: perceptron, multilayer perceptron, radial basis network, probabilistic neu... 4.8.2 Backpropagation neural networks 4.9 Discriminant analysis family models 4.9.1 Family models: linear discriminant analysis, quadratic discriminant analysis 4.10 Logistic regression classifier 4.10.1 Family models “logistic regression” 4.11 Ensemble classifiers family models 4.11.1 Models: AdaBoost, RUSBoost, Subspace kNN, Random Forrest, Subspace discriminant 4.12 IBM ML Solution: IBM Watson SPSS 4.12.1 SPSS Modeler flows %3e Modeling 4.12.2 SPSS Modeler flows %3e Output 4.12.2.1 Research 4.1 Tutorial IBM Watson SPSS Modeler Flow for “ML Model for Diabetes” 4.12.2.1.1 Case for research 4.12.2.1.2 General objective 4.12.2.1.3 Specific objectives 4.12.2.1.4 Dataset 4.12.2.1.5 Background for “Diabetes” 4.12.2.1.6 Procedure 4.12.2.2 Research tutorial 4.2 IBM Watson SPSS Modeler Flow for “Heart disease ML model and deployment” 4.12.2.2.1 Case for research 4.12.2.2.2 General objective 4.12.2.2.3 Background for “Heart diseases” 4.12.2.2.4 Specific objectives 4.12.2.2.5 Dataset 4.12.2.2.6 Procedure 4.12.2.3 Research tutorial 4.3 IBM Watson SPSS Modeler Flow for “Kidney disease ML Auto Classifiers Models and deploy the b... 4.12.2.3.1 Case for research 4.12.2.3.2 General objective for this research 4.12.2.3.3 Background for “Kidney disease” 4.12.2.3.4 Specific objectives 4.12.2.3.5 Dataset 4.12.2.3.6 Procedure 4.12.2.4 Research tutorial 4.4 IBM Watson AutoAI experimenter for “Breast cancer ML model and deploy the best model” 4.12.2.4.1 Case for research 4.12.2.4.2 General objective 4.12.2.4.3 Background for “Breast cancer” 4.12.2.4.4 Specific objectives 4.12.2.4.5 Dataset 4.12.2.4.6 Procedure 4.12.2.5 Research tutorial 4.5 MATLAB: Statistics and Machine Learning Toolbox for a “Diabetes dataset AI modeling for Clas... 4.12.2.5.1 Case for research 4.12.2.5.2 General objective for this research 4.12.2.5.3 Specific objectives 4.12.2.5.4 Dataset 4.12.2.5.5 Procedure References Further reading 5 Deep Learning Models Principles Applied to Biomedical Engineering 5.1 Deep learning based on artificial neural networks 5.2 Feed forward neural networks types 5.2.1 Perceptron (P) or single-layer perceptron network 5.2.1.1 ANN activation functions 5.2.2 Multilayer perceptron 5.2.3 Radial basis function network 5.2.4 Probabilistic neural network 5.2.5 Extreme Learning Machine 5.3 Shallow neural network 5.3.1 Research 5.1 Feed Forward Neural Network to Analyze “Human Body Fat” 5.3.1.1 Case for research 5.3.1.2 General objective 5.3.1.3 Specific objectives 5.3.1.4 Background for “Human Body Fat” 5.3.1.5 Dataset 5.3.1.6 Procedure 5.3.2 Research 52 Neural Network for clustering based on “Self-Organizing MAP* through a Shallow Neural Network” to analyze... 5.3.2.1 Case for research 5.3.2.2 General objective 5.3.2.3 Specific objectives 5.3.2.4 Background for “sEMG” 5.3.2.5 Background for “Diabetes Mellitus” 5.3.2.6 Dataset 5.3.2.7 Procedure 5.3.3 Research 5.3 Neural Network for Dynamic Time series based on a “NARX is a nonlinear autoregressive exogenous model” t... 5.3.3.1 Case for research 5.3.3.2 General objective 5.3.3.3 Specific objectives 5.3.3.4 Background for “vertical Ground Reaction Forces” 5.3.3.5 Dataset 5.3.3.6 Procedure 5.4 Backpropagation neural networks types 5.4.1 Auto Encoder 5.4.2 Variational Auto Encoder 5.4.3 Denoising Auto Encoder 5.4.4 Sparse Auto Encoder and stacked auto encoders 5.4.4.1 Research 5.4* Backpropagation Neural Network for Patterns Recognition and classification of “Breast Cancer” 5.4.4.1.1 Case for research 5.4.4.1.2 General objective 5.4.4.1.3 Specific objectives 5.4.4.1.4 Background for “Breast cancer” 5.4.4.1.5 Dataset 5.4.4.1.6 Procedure 5.4.5 Deep Convolution Network or ConvNet 5.4.6 Deconvolutional network 5.4.7 Deep Convolutional Inverse Graphics Network 5.4.8 Generative Adversarial Network 5.4.9 Deep Residual Network or Deep ResNet 5.5 Transfer learning from pretrained deep learning networks 5.5.1 Research 5.5 “Pretrained Deep Convolutional Neural Network to obtain an AI model to classify Mammograms standard view... 5.5.1.1 Case for research 5.5.1.2 General objective 5.5.1.3 Specific objectives 5.5.1.4 Background for “Mammography views” 5.5.1.5 Dataset 5.5.1.6 Procedure 5.5.2 Research 5.6 modify a “Pretrained Deep Convolutional Neural Network” to obtain an AI model to “classify Mammograms vi... 5.5.2.1 Case for research 5.5.2.2 General objective 5.5.2.3 Background “Mammograms to detect breast abnormalities” 5.5.2.4 Specific objectives 5.5.2.5 Dataset 5.5.2.6 Procedure 5.5.3 Research 5.7 “custom Deep Convolutional Neural Network” to obtain an AI model to “classify Cervical X-rays view types” 5.5.3.1 Case for research 5.5.3.2 General objective 5.5.3.3 Background 5.5.3.4 Specific objectives 5.5.3.5 Dataset 5.5.3.6 Procedure References 6 Deep Learning Models Evolution Applied to Biomedical Engineering 6.1 Deep learning models evolution 6.2 Recurrent neural networks types 6.2.1 Recurrent Neural Network vanilla 6.2.2 Long/short-term memory 6.2.3 Gated recurrent unit networks 6.2.3.1 Research 6.1 LSTM to classify videos about human body movements and detect human falls 6.2.3.1.1 Case for research 6.2.3.1.2 General objective 6.2.3.1.3 Specific objectives 6.2.3.1.4 Background for “Walking Human falls” 6.2.3.1.5 Dataset 6.2.3.1.6 Procedure 6.2.4 Recurrent convolutional neural networks 6.2.5 Regional-Convolutional Neural Network Object detection in AI models 6.2.5.1 Research 6.2 Regional-CNN model for object detection of breast tumor in mammogram 6.2.5.1.1 Case for research 6.2.5.1.2 General objective 6.2.5.1.3 Specific objectives 6.2.5.1.4 Background for “breast tumors in mammogram” 6.2.5.1.5 Dataset 6.2.5.1.6 Procedure 6.2.6 Hopfield Network 6.2.6.1 Research 6.3 Hopfield Network model to reconstruct noisy chest X-ray images 6.2.6.1.1 Case for research 6.2.6.1.2 General objective 6.2.6.1.3 Specific objectives 6.2.6.1.4 Background for “chest X-ray images” 6.2.6.1.5 Dataset 6.2.6.1.6 Procedure 6.2.7 Boltzmann Machine 6.2.8 Restricted Boltzmann Machine 6.2.8.1 Research 6.4 Restricted Boltzmann Machine model to reconstruct noisy chest X-ray images 6.2.8.1.1 Case for research 6.2.8.1.2 General objective 6.2.8.1.3 Specific objectives 6.2.8.1.4 Background for “chest X-ray images” 6.2.8.1.5 Dataset 6.2.8.1.6 Procedure 6.2.9 Liquid State Machine 6.2.10 Echo State Network 6.2.10.1 Research 6.5 Create a Reservoir Computing approach for a simulation of “Liquid State Machine (LSM)” of node-neuron... 6.2.10.1.1 Case for research 6.2.10.1.2 General objective 6.2.10.1.3 Specific objectives 6.2.10.1.4 Background for “Izhikevich neuronal mathematical model” 6.2.10.1.5 Background for “COVID-19 lung imaging chest X-rays” 6.2.10.1.6 Dataset 6.2.10.1.7 Procedure 6.2.11 Korhonen Network 6.3 Memory augmented neural networks types 6.3.1 Neural Turing machine 6.3.2 Differentiable Neural Computers 6.3.2.1 Research 6.6 simulation of a Turing Machine (TM) using a recursive function 6.3.2.1.1 Case for research 6.3.2.1.2 General objective 6.3.2.1.3 Specific objectives 6.3.2.1.4 Background for “Modes of transmission of SARS-CoV-2 (COVID-19)” 6.3.2.1.5 Dataset 6.3.2.1.6 Procedure 6.4 Modular Neural Networks types 6.4.1 Deep Belief Network 6.4.1.1 Research 6.7 Create a Deep Belief Network model to analyze and differentiate normal and pneumonia chest X-rays 6.4.1.1.1 Case for research 6.4.1.1.2 General objective 6.4.1.1.3 Specific objectives 6.4.1.1.4 Background for “COVID-19 lung imaging chest X-rays 6.4.1.1.5 Dataset 6.4.1.1.6 Procedure 6.5 Evolutionary Deep Neural Networks types 6.5.1 Capsule Networks 6.5.2 Attention networks References Further reading 7 Cognitive learning and reasoning models applied to biomedical engineering 7.1 Introduction 7.2 Artificial intelligence and Cognitive Computing Agents System (AI-CCAS) 7.3 Inference engine and research example 7.3.1 Research 7.1 7.3.2 Research 7.2 7.4 Action generation 7.5 Business intelligence in healthcare 7.6 Learning and reasoning relationship of biomedical engineering, cognitive science, and computer science through artifici... 7.6.1 Cognitive learning and reasoning 7.6.2 Deductive reasoning 7.6.2.1 “Deductive reasoning—Propositional Arguments—Modus Ponens” 7.6.2.2 “Deductive reasoning—Propositional Arguments—Modus Tollens” 7.6.2.3 “Deductive reasoning—Propositional Arguments—Disjunction Elimination” 7.6.3 Inductive reasoning 7.6.3.1 “Inductive reasoning—Generalized arguments” 7.6.3.2 “Inductive reasoning—Statistical arguments” 7.6.3.3 “Inductive reasoning—Analogical arguments” 7.6.3.4 “Inductive reasoning—Predictive arguments” 7.6.3.5 “Inductive reasoning—Causal arguments” 7.6.4 Abductive reasoning 7.6.4.1 “Abductive reasoning—Causal arguments—Inference of abduction” 7.6.4.1.1 Example: “Inductive reasoning—Causal arguments—Inference of abduction” 7.6.4.2 “Abductive reasoning—Causal arguments—Backward-chaining” 7.6.4.2.1 Example “Abductive reasoning—Causal arguments—Backward-chaining” 7.6.4.3 “Abductive reasoning—Causal arguments—Paradigm case-base” 7.6.4.3.1 Example: “Abductive reasoning —Causal arguments—Paradigm case-based” 7.6.4.4 “Abductive reasoning—Causal arguments—Generative coherence metric (Human inference)” 7.6.4.4.1 Example: “Abductive Reasoning—Causal arguments—Generative Coherence metric” 7.6.5 Abductive reasoning for medical diagnosis 7.6.5.1 Example: Abductive reasoning for medical diagnosis 7.6.6 Metaphoric reasoning 7.6.6.1 Example: metaphoric reasoning 7.6.7 Neuro-Fuzzy logic reasoning as cognitive reasoning 7.6.7.1 Example “Fuzzy System: Mamdani-type inference” in MATLAB 7.6.7.2 Example “Fuzzy System: Sugeno-type inference” in MATLAB 7.6.7.3 “Neuro-Fuzzy systems (NFS)” 7.6.8 Visuospatial relational reasoning 7.6.9 Cognitive learning and relationship with neuroscience of reasoning 7.7 Cognitive Learning and Reasoning research example applying AI-CCAS framework 7.7.1 Research 7.3 7.7.2 Research 7.4 7.7.3 Research 7.5 7.8 Challenge research for “Applied Biomedical Engineering using Artificial Intelligence and Cognitive Models” 7.8.1 Challenge research project # 1: “Inductive Reasoning AI evaluation test for neurologic diseases patients under Co... 7.8.2 Challenge research project # 2: “Abducting Reasoning using AI evaluation tests for patients under Cognitive Learn... 7.8.3 Challenge research project # 3: “Metaphoric reasoning for clinical diagnosis using Cognitive Learning and Reasoni... 7.8.4 Challenge research project # 4 “Neurologic – evaluating anxiety in neurologic diseases using Cognitive Therapy Th... 7.8.5 Challenge research project # 5: Analyze Neurologic opinion words with positive and negative frequently used to de... 7.8.6 Challenge research project # 6: “Classify status of neurologic disease patients analyzing their images, movements... 7.8.7 Challenge research project # 7: “human voice cognitive analysis for cognitive services as voice therapy” 7.8.8 Challenge research project # 8: Cognitive Behavioral Therapy 7.8.9 Challenge research project # 9: detection of “prefatigue/fatigue by stress and anxiety” 7.8.10 Top Challenge research project # 10 building a Cognitive health Dashboard References Further Reading Index Back Cover