Marianne Sarazin is a public health doctor with a doctorate in Life Sciences from the Engineering and Health Center (CIS) of the École nationale supérieure des mines de Saint-Étienne, France. She is the Head of the Medical Information department of the Mutualiste Sanitary Group of Saint-Étienne (Aesio group) and a CIS collaborator in the Optimization of Healthcare Systems department as well as UMRS 1136 Inserm, specializing in the modeling of epidemics.

• Author Presentation xiPreface xvMarianne SARAZINPart 1 Growing with Artificial Intelligence 1Introduction to Part 1 3Marianne SARAZINChapter 1 From Human to Artificial Intelligence 5Bruno SALGUES1.1 The different forms of intelligence 51.1.1 Human intelligence typologies 51.1.2 Artificial intelligence (AI) and human intelligence 51.1.3 Object intelligence and human assistance 51.2 History of "artificial" intelligence 71.2.1 Mechanical forms 71.2.2 The desire to model neurons and cybernetics 81.2.3 The arrival of computers 91.2.4 Different uses of artificial intelligence in healthcare 121.2.5 Automated fields or scored answers 151.2.6 Expert systems 161.2.7 Vector differentiation or vector forest 171.2.8 Convolution matrix 181.2.9 Multi-cameral or democratic systems 191.2.10 System dynamics 191.2.11 Machine learning 201.2.12 Deep learning 221.2.13 The keys to adopting artificial intelligence in healthcare 231.2.14 Organ processing using artificial intelligence 241.2.15 Medical procedures aided by artificial intelligence: drug dosage 25Chapter 2 The Philosopher’s Point of View: The Challenges of AI for Our Humanity 29François-Xavier CLÉMENT2.1 Introduction 292.2 The beginnings of AI 302.3 Man: master or slave of AI – examples of behavioral approaches 332.3.1 Teenagers and their phones 332.3.2 Consumer behavior and AI 342.3.3 Memory use and AI 342.3.4 Human intelligence versus AI 352.3.5 Game master: human or AI? 372.4 And what about humanity? 382.4.1 New language 382.4.2 New thinking 402.4.3 A new moral 412.5 AI in education: changing learning styles among students in 2020 432.5.1 Generations and technology 432.5.2 Student behavior 442.6 A few concluding words 45Part 2 Working with Artificial Intelligence 47Introduction to Part 2 49Marianne SARAZINChapter 3 For Strategic and Responsible "Piloting" of AI-related Open Innovation Projects 51Aline COURIE-LEMEUR3.1 Introduction 513.2 Innovation development and the "open innovation" model: challenges and risks 523.2.1 Definition of open innovation 523.2.2 Dangers of open innovation 543.3 Steering "open innovation" projects as part of the development of artificial intelligence techniques 563.3.1 Strategic and responsible management 573.3.2 The attributes of strategic, responsible management 603.4 In a nutshell 633.5 Conclusion 66Chapter 4 Management and AI: Myths and Realities 67Gilles ROUET4.1 Introduction 674.1.1 Paradigm shift in the business world 674.1.2 New management model 684.1.3 Mixing genres: intrusion of everyday technologies into the business world 684.2 Management in our digital environment 694.2.1 The contribution of digital technology to companies 704.2.2 Artificial intelligence 724.2.3 Analysis concepts for large databases 734.3 Humans and machines 764.3.1 Optimizing AI-based tools within companies 774.3.2 Putting people before AI in companies 784.3.3 Evolution of essential skills 794.3.4 And then… 804.4 Conclusion 81Part 3 Managing Healthcare with Artificial Intelligence 83Introduction to Part 3 85Marianne SARAZINChapter 5 How to Bring the Medical World Out of the Pre-digital Age? 87Marc SOLER5.1 Introduction 875.2 Healthcare professionals’ relationship with digital technology 885.2.1 Level of training of healthcare professionals 885.2.2 Technology and healthcare professionals 905.3 Creating a universal medical record: a utopia? 935.4 "Artificial intelligence" at the service of healthcare: the French government’s position 955.5 The approach of technology suppliers? 965.6 IBM’s Watson system: its history and application to the medical field 985.6.1 Concept 985.6.2 Oncology applications 995.6.3 An admission of failure 995.6.4 The clinician’s point of view 1005.6.5 Illustrative examples 1025.6.6 The importance of source data 1045.6.7 Conclusion 1065.7 The role of start-ups 1075.7.1 A few examples 1085.7.2 Special case of BenevolentAI 1095.8 Conclusion 111Chapter 6 Data Quality: A Major Challenge for AI in Healthcare 113Marysa GERMAIN6.1 Introduction 1136.2 From patient data to AI 1146.2.1 The legal framework for processing medical data 1146.2.2 The technological framework 1176.2.3 The hospital’s database management department: the DIM 1186.2.4 Caregivers’ views on the digitization of medical information 1206.3 Data quality and consolidation 1206.3.1 Medical data 1206.3.2 Implementation of a continuous process of quality improvement 122Part 4 Aging with Artificial Intelligence 133Introduction to Part 4 135Marianne SARAZINChapter 7 Proposed Method for Developing an Aging Score 137Marianne SARAZIN7.1 Introduction 1377.2 Focus on the determinants of age-related frailty 1387.3 Choice of marker variables to determine age 1397.3.1 Rational marker selection based on expertise: an operational approach based on a literature review 1397.3.2 Selection of markers using variables with values within normality limits 1417.3.3 Conclusion 1447.4 Choice of normal aging control population 1457.4.1 Initial hypotheses defining the choice of the control population and the construction of the score 1457.4.2 First approach: rational selection of the control population based on the literature 1467.4.3 Second approach: selection of the control population by classification using the dynamic clustering method 1467.4.4 Results 1487.4.5 Conclusion 1507.5 Mathematical modeling of the aging score 1517.5.1 Initial concept 1517.5.2 Calculating biological age from a control population sample 1527.5.3 Conclusions 1597.6 Calculating biological age: modeling dependence between marker variables using a Gaussian copula 1607.6.1 Method 1607.6.2 Source population 1627.6.3 Results 1627.6.4 Conclusions 1667.7 Calculation of biological age for any population (using a Gaussian copula) 1667.7.1 Method 1667.7.2 Source population 1697.7.3 Results 1697.7.4 Conclusions 1757.8 Perspectives on this work 1767.8.1 Advantages and limitations of this work 1767.8.2 Perspectives 178Chapter 8 Automatic Detection of Behavioral Changes in a Smart Home 179Cyriak AZEFAC8.1 Introduction 1798.2 Definitions 1818.3 Methodology 1828.3.1 Attribute extraction 1848.3.2 Unsupervised classification 1868.3.3 Auto-encoder 1878.3.4 Clustering 1888.4 Case study: ARUBA 1888.5 Conclusion 189Conclusion 191Marianne SARAZINReferences 193List of Authors 203Index 205

Marianne Sarazin is a public health doctor with a doctorate in Life Sciences from the Engineering and Health Center (CIS) of the École nationale supérieure des mines de Saint-Étienne, France. She is the Head of the Medical Information department of the Mutualiste Sanitary Group of Saint-Étienne (Aesio group) and a CIS collaborator in the Optimization of Healthcare Systems department as well as UMRS 1136 Inserm, specializing in the modeling of epidemics.

• Author Presentation xiPreface xvMarianne SARAZINPart 1 Growing with Artificial Intelligence 1Introduction to Part 1 3Marianne SARAZINChapter 1 From Human to Artificial Intelligence 5Bruno SALGUES1.1 The different forms of intelligence 51.1.1 Human intelligence typologies 51.1.2 Artificial intelligence (AI) and human intelligence 51.1.3 Object intelligence and human assistance 51.2 History of "artificial" intelligence 71.2.1 Mechanical forms 71.2.2 The desire to model neurons and cybernetics 81.2.3 The arrival of computers 91.2.4 Different uses of artificial intelligence in healthcare 121.2.5 Automated fields or scored answers 151.2.6 Expert systems 161.2.7 Vector differentiation or vector forest 171.2.8 Convolution matrix 181.2.9 Multi-cameral or democratic systems 191.2.10 System dynamics 191.2.11 Machine learning 201.2.12 Deep learning 221.2.13 The keys to adopting artificial intelligence in healthcare 231.2.14 Organ processing using artificial intelligence 241.2.15 Medical procedures aided by artificial intelligence: drug dosage 25Chapter 2 The Philosopher’s Point of View: The Challenges of AI for Our Humanity 29François-Xavier CLÉMENT2.1 Introduction 292.2 The beginnings of AI 302.3 Man: master or slave of AI – examples of behavioral approaches 332.3.1 Teenagers and their phones 332.3.2 Consumer behavior and AI 342.3.3 Memory use and AI 342.3.4 Human intelligence versus AI 352.3.5 Game master: human or AI? 372.4 And what about humanity? 382.4.1 New language 382.4.2 New thinking 402.4.3 A new moral 412.5 AI in education: changing learning styles among students in 2020 432.5.1 Generations and technology 432.5.2 Student behavior 442.6 A few concluding words 45Part 2 Working with Artificial Intelligence 47Introduction to Part 2 49Marianne SARAZINChapter 3 For Strategic and Responsible "Piloting" of AI-related Open Innovation Projects 51Aline COURIE-LEMEUR3.1 Introduction 513.2 Innovation development and the "open innovation" model: challenges and risks 523.2.1 Definition of open innovation 523.2.2 Dangers of open innovation 543.3 Steering "open innovation" projects as part of the development of artificial intelligence techniques 563.3.1 Strategic and responsible management 573.3.2 The attributes of strategic, responsible management 603.4 In a nutshell 633.5 Conclusion 66Chapter 4 Management and AI: Myths and Realities 67Gilles ROUET4.1 Introduction 674.1.1 Paradigm shift in the business world 674.1.2 New management model 684.1.3 Mixing genres: intrusion of everyday technologies into the business world 684.2 Management in our digital environment 694.2.1 The contribution of digital technology to companies 704.2.2 Artificial intelligence 724.2.3 Analysis concepts for large databases 734.3 Humans and machines 764.3.1 Optimizing AI-based tools within companies 774.3.2 Putting people before AI in companies 784.3.3 Evolution of essential skills 794.3.4 And then… 804.4 Conclusion 81Part 3 Managing Healthcare with Artificial Intelligence 83Introduction to Part 3 85Marianne SARAZINChapter 5 How to Bring the Medical World Out of the Pre-digital Age? 87Marc SOLER5.1 Introduction 875.2 Healthcare professionals’ relationship with digital technology 885.2.1 Level of training of healthcare professionals 885.2.2 Technology and healthcare professionals 905.3 Creating a universal medical record: a utopia? 935.4 "Artificial intelligence" at the service of healthcare: the French government’s position 955.5 The approach of technology suppliers? 965.6 IBM’s Watson system: its history and application to the medical field 985.6.1 Concept 985.6.2 Oncology applications 995.6.3 An admission of failure 995.6.4 The clinician’s point of view 1005.6.5 Illustrative examples 1025.6.6 The importance of source data 1045.6.7 Conclusion 1065.7 The role of start-ups 1075.7.1 A few examples 1085.7.2 Special case of BenevolentAI 1095.8 Conclusion 111Chapter 6 Data Quality: A Major Challenge for AI in Healthcare 113Marysa GERMAIN6.1 Introduction 1136.2 From patient data to AI 1146.2.1 The legal framework for processing medical data 1146.2.2 The technological framework 1176.2.3 The hospital’s database management department: the DIM 1186.2.4 Caregivers’ views on the digitization of medical information 1206.3 Data quality and consolidation 1206.3.1 Medical data 1206.3.2 Implementation of a continuous process of quality improvement 122Part 4 Aging with Artificial Intelligence 133Introduction to Part 4 135Marianne SARAZINChapter 7 Proposed Method for Developing an Aging Score 137Marianne SARAZIN7.1 Introduction 1377.2 Focus on the determinants of age-related frailty 1387.3 Choice of marker variables to determine age 1397.3.1 Rational marker selection based on expertise: an operational approach based on a literature review 1397.3.2 Selection of markers using variables with values within normality limits 1417.3.3 Conclusion 1447.4 Choice of normal aging control population 1457.4.1 Initial hypotheses defining the choice of the control population and the construction of the score 1457.4.2 First approach: rational selection of the control population based on the literature 1467.4.3 Second approach: selection of the control population by classification using the dynamic clustering method 1467.4.4 Results 1487.4.5 Conclusion 1507.5 Mathematical modeling of the aging score 1517.5.1 Initial concept 1517.5.2 Calculating biological age from a control population sample 1527.5.3 Conclusions 1597.6 Calculating biological age: modeling dependence between marker variables using a Gaussian copula 1607.6.1 Method 1607.6.2 Source population 1627.6.3 Results 1627.6.4 Conclusions 1667.7 Calculation of biological age for any population (using a Gaussian copula) 1667.7.1 Method 1667.7.2 Source population 1697.7.3 Results 1697.7.4 Conclusions 1757.8 Perspectives on this work 1767.8.1 Advantages and limitations of this work 1767.8.2 Perspectives 178Chapter 8 Automatic Detection of Behavioral Changes in a Smart Home 179Cyriak AZEFAC8.1 Introduction 1798.2 Definitions 1818.3 Methodology 1828.3.1 Attribute extraction 1848.3.2 Unsupervised classification 1868.3.3 Auto-encoder 1878.3.4 Clustering 1888.4 Case study: ARUBA 1888.5 Conclusion 189Conclusion 191Marianne SARAZINReferences 193List of Authors 203Index 205

Marianne Sarazin is a public health doctor with a doctorate in Life Sciences from the Engineering and Health Center (CIS) of the École nationale supérieure des mines de Saint-Étienne, France. She is the Head of the Medical Information department of the Mutualiste Sanitary Group of Saint-Étienne (Aesio group) and a CIS collaborator in the Optimization of Healthcare Systems department as well as UMRS 1136 Inserm, specializing in the modeling of epidemics.

• Author Presentation xiPreface xvMarianne SARAZINPart 1 Growing with Artificial Intelligence 1Introduction to Part 1 3Marianne SARAZINChapter 1 From Human to Artificial Intelligence 5Bruno SALGUES1.1 The different forms of intelligence 51.1.1 Human intelligence typologies 51.1.2 Artificial intelligence (AI) and human intelligence 51.1.3 Object intelligence and human assistance 51.2 History of "artificial" intelligence 71.2.1 Mechanical forms 71.2.2 The desire to model neurons and cybernetics 81.2.3 The arrival of computers 91.2.4 Different uses of artificial intelligence in healthcare 121.2.5 Automated fields or scored answers 151.2.6 Expert systems 161.2.7 Vector differentiation or vector forest 171.2.8 Convolution matrix 181.2.9 Multi-cameral or democratic systems 191.2.10 System dynamics 191.2.11 Machine learning 201.2.12 Deep learning 221.2.13 The keys to adopting artificial intelligence in healthcare 231.2.14 Organ processing using artificial intelligence 241.2.15 Medical procedures aided by artificial intelligence: drug dosage 25Chapter 2 The Philosopher’s Point of View: The Challenges of AI for Our Humanity 29François-Xavier CLÉMENT2.1 Introduction 292.2 The beginnings of AI 302.3 Man: master or slave of AI – examples of behavioral approaches 332.3.1 Teenagers and their phones 332.3.2 Consumer behavior and AI 342.3.3 Memory use and AI 342.3.4 Human intelligence versus AI 352.3.5 Game master: human or AI? 372.4 And what about humanity? 382.4.1 New language 382.4.2 New thinking 402.4.3 A new moral 412.5 AI in education: changing learning styles among students in 2020 432.5.1 Generations and technology 432.5.2 Student behavior 442.6 A few concluding words 45Part 2 Working with Artificial Intelligence 47Introduction to Part 2 49Marianne SARAZINChapter 3 For Strategic and Responsible "Piloting" of AI-related Open Innovation Projects 51Aline COURIE-LEMEUR3.1 Introduction 513.2 Innovation development and the "open innovation" model: challenges and risks 523.2.1 Definition of open innovation 523.2.2 Dangers of open innovation 543.3 Steering "open innovation" projects as part of the development of artificial intelligence techniques 563.3.1 Strategic and responsible management 573.3.2 The attributes of strategic, responsible management 603.4 In a nutshell 633.5 Conclusion 66Chapter 4 Management and AI: Myths and Realities 67Gilles ROUET4.1 Introduction 674.1.1 Paradigm shift in the business world 674.1.2 New management model 684.1.3 Mixing genres: intrusion of everyday technologies into the business world 684.2 Management in our digital environment 694.2.1 The contribution of digital technology to companies 704.2.2 Artificial intelligence 724.2.3 Analysis concepts for large databases 734.3 Humans and machines 764.3.1 Optimizing AI-based tools within companies 774.3.2 Putting people before AI in companies 784.3.3 Evolution of essential skills 794.3.4 And then… 804.4 Conclusion 81Part 3 Managing Healthcare with Artificial Intelligence 83Introduction to Part 3 85Marianne SARAZINChapter 5 How to Bring the Medical World Out of the Pre-digital Age? 87Marc SOLER5.1 Introduction 875.2 Healthcare professionals’ relationship with digital technology 885.2.1 Level of training of healthcare professionals 885.2.2 Technology and healthcare professionals 905.3 Creating a universal medical record: a utopia? 935.4 "Artificial intelligence" at the service of healthcare: the French government’s position 955.5 The approach of technology suppliers? 965.6 IBM’s Watson system: its history and application to the medical field 985.6.1 Concept 985.6.2 Oncology applications 995.6.3 An admission of failure 995.6.4 The clinician’s point of view 1005.6.5 Illustrative examples 1025.6.6 The importance of source data 1045.6.7 Conclusion 1065.7 The role of start-ups 1075.7.1 A few examples 1085.7.2 Special case of BenevolentAI 1095.8 Conclusion 111Chapter 6 Data Quality: A Major Challenge for AI in Healthcare 113Marysa GERMAIN6.1 Introduction 1136.2 From patient data to AI 1146.2.1 The legal framework for processing medical data 1146.2.2 The technological framework 1176.2.3 The hospital’s database management department: the DIM 1186.2.4 Caregivers’ views on the digitization of medical information 1206.3 Data quality and consolidation 1206.3.1 Medical data 1206.3.2 Implementation of a continuous process of quality improvement 122Part 4 Aging with Artificial Intelligence 133Introduction to Part 4 135Marianne SARAZINChapter 7 Proposed Method for Developing an Aging Score 137Marianne SARAZIN7.1 Introduction 1377.2 Focus on the determinants of age-related frailty 1387.3 Choice of marker variables to determine age 1397.3.1 Rational marker selection based on expertise: an operational approach based on a literature review 1397.3.2 Selection of markers using variables with values within normality limits 1417.3.3 Conclusion 1447.4 Choice of normal aging control population 1457.4.1 Initial hypotheses defining the choice of the control population and the construction of the score 1457.4.2 First approach: rational selection of the control population based on the literature 1467.4.3 Second approach: selection of the control population by classification using the dynamic clustering method 1467.4.4 Results 1487.4.5 Conclusion 1507.5 Mathematical modeling of the aging score 1517.5.1 Initial concept 1517.5.2 Calculating biological age from a control population sample 1527.5.3 Conclusions 1597.6 Calculating biological age: modeling dependence between marker variables using a Gaussian copula 1607.6.1 Method 1607.6.2 Source population 1627.6.3 Results 1627.6.4 Conclusions 1667.7 Calculation of biological age for any population (using a Gaussian copula) 1667.7.1 Method 1667.7.2 Source population 1697.7.3 Results 1697.7.4 Conclusions 1757.8 Perspectives on this work 1767.8.1 Advantages and limitations of this work 1767.8.2 Perspectives 178Chapter 8 Automatic Detection of Behavioral Changes in a Smart Home 179Cyriak AZEFAC8.1 Introduction 1798.2 Definitions 1818.3 Methodology 1828.3.1 Attribute extraction 1848.3.2 Unsupervised classification 1868.3.3 Auto-encoder 1878.3.4 Clustering 1888.4 Case study: ARUBA 1888.5 Conclusion 189Conclusion 191Marianne SARAZINReferences 193List of Authors 203Index 205