Brain-Machine Interface : Closed-loop Bidirectional System Design

This book provides an introduction to the emerging area of "Brain-Machine Interfaces," with emphasis on the operation and practical design aspects. The book will help both electrical & bioengineers as well as neuroscience investigators to learn about the next generation brain-machine interfaces. The comprehensive review and design analysis will be very helpful for researchers who are new to this area or interested in the study of the brain. The in-depth discussion of practical design issues especially in animal experiments will also be valuable for experienced researchers. Provides readers with background on brain-machine interfaces, a summary of design requirements, and circuit and system design examples ; Presents the first comprehensive study of closed-loop bidirectional BMI's for freely behaving animals ; Demonstrates the first reported portable system to provide all necessary hardware for a closed-loop sensorimotor neural interf ace ; many innovations in the circuit design optimized for the characteristics of neural interface and electrode-electrolyte interface Preface 5 Contents 7 About the Authors 10 Acronyms 12 List of Figures 13 List of Tables 30 1 Introduction 31 1.1 Background and Motivation 31 1.2 Review of Prior Work 34 1.3 Overview of the Bidirectional Closed-Loop Brain–Machine Interface System 40 1.4 Outline of This Book 43 2 Neural Recording Front-End Design 46 2.1 Introduction 46 2.1.1 Signal Characteristics 47 2.1.2 Design Specifications 48 2.2 Design of a Low-Noise Neural Amplifier 50 2.2.1 Review of Prior Work 50 2.2.1.1 System Topology 50 2.2.1.2 Low-Noise OTA 52 2.2.1.3 Other Noise Reducing Techniques 52 2.2.2 Circuit Implementation 53 2.2.3 Measurement Results 58 2.3 A Pre-whitening Neural Amplifier 61 2.3.1 Introduction 61 2.3.2 Analysis of Pre-whitening Neural Amplifier Design 62 2.3.2.1 Pre-whitening After a Wideband Low-Noise Amplifier 64 2.3.2.2 Low-Noise Neural Amplifier with Integrated Pre-whitening Filter 66 2.3.3 Circuit Implementation 69 2.3.4 Measurement Results 70 2.4 Design of a Low-Power Analog-to-Digital Converter 74 2.4.1 Introduction 74 2.4.2 Circuit Implementation 75 2.4.3 Measurement Results 77 2.5 A Compressed Sensing Neural Signal Acquisition System 80 2.5.1 Introduction 80 2.5.2 A Brief Background of Compressed Sensing 82 2.5.2.1 Compression Process 82 2.5.2.2 Reconstruction Process 82 2.5.2.3 Reconstruction Evaluation Criteria 83 2.5.3 System Overview 83 2.5.4 Circuit Implementation 85 2.5.4.1 Energy Efficient Analog Front-End 85 2.5.4.2 Compressed Sensing Module 88 2.5.4.3 On-Chip Wireless Power and Data Link 90 2.5.4.4 External Wireless Relay Board 91 2.5.5 Measurement Results 92 3 Neural Feature Extraction 98 3.1 Introduction 98 3.2 Natural Logarithmic Domain Field Potential Energy Extraction 100 3.2.1 Introduction 100 3.2.2 System and Circuits Implementation 101 3.2.2.1 Design of the GmC Filter 101 3.2.2.2 Biasing Current Generation 104 3.2.2.3 Design of the Biquad Filter 106 3.2.2.4 Multiplier and Integrator 107 3.2.3 Measurement Results 108 3.3 Action Potential Discrimination 111 3.3.1 Introduction 111 3.3.1.1 Integrate and Fire Model 112 3.3.1.2 Review of Action Potential Discrimination Methods 113 3.3.2 Circuit Implementation 115 3.3.3 Experimental Results 118 3.4 Matched Filter for Neural Feature Extraction 121 3.4.1 Introduction 121 3.4.2 Matched Filter and Pre-whitening for Optimum Correlation Detection 122 3.4.3 Methodologies 124 3.4.3.1 Dataset 124 3.4.3.2 Bandpass Filter 125 3.4.3.3 Matched Filter with Pre-whitening 126 3.4.4 Experimental Results 127 3.4.4.1 Detection of Synthesized Signal 127 3.4.4.2 Detection of Recorded Neural Signal 128 3.4.4.3 Compressive Domain Matched Filtering 129 4 Neural Stimulator Design 132 4.1 Introduction 132 4.1.1 Background of Neurostimulation 133 4.1.2 Electrode and Electrolyte Interface 133 4.2 Overview of Electrical Stimulator Design 134 4.2.1 Methods of Stimuli Generation 136 4.2.1.1 Voltage-Regulated Stimulation 136 4.2.1.2 Current-Regulated Stimulation 137 4.2.1.3 Charges-Regulated Stimulation 137 4.2.2 Stimulation Waveform and Electrode Configuration 137 4.2.3 Methods for Charge Balancing 139 4.2.3.1 Matching and Calibration 139 4.2.3.2 Passive and Active Discharge 140 4.3 Design of a General–Purpose Stimulator 141 4.3.1 Architecture of the Stimulator 141 4.3.2 Circuit Implementation 142 4.3.3 Measurement Results 144 4.4 An Energy Efficient Net-Zero Charge Neural Stimulator 147 4.4.1 Introduction 147 4.4.2 Motivation and Innovation 148 4.4.2.1 Net-Zero Charges Stimulation 148 4.4.2.2 Adaptive Driving Voltage 151 4.4.2.3 Arbitrary Channel Configuration 152 4.4.3 Circuit Implementation 152 4.4.3.1 System Architecture 152 4.4.3.2 Output Stage with Dynamic Element Matching 154 4.4.3.3 Feed-Forward Error Compensation Comparator 156 4.4.4 Experimental Results 159 4.4.5 Conclusion 163 5 Bidirectional Neural Interface and Closed-Loop Control 165 5.1 Introduction 165 5.2 Stimulation Artifacts in the Bidirectional Neural Interface 166 5.2.1 Introduction 166 5.2.2 Review of Prior Work 167 5.2.3 Analysis of Stimulation Artifacts 168 5.2.3.1 Origins of Stimulation Artifacts 168 5.2.3.2 Configuration of the Interface Circuits 170 5.2.3.3 Practical Issues in Circuit Design 172 5.2.4 Methods 173 5.2.5 Experimental Results 175 5.2.5.1 In Vitro Experiment 175 5.2.5.2 In Vivo Experiment 178 5.2.6 Conclusion 180 5.3 Closed-Loop Neural Interface System 183 5.3.1 Introduction 183 5.3.2 Mechanism of Closed-Loop Neural Interface System 184 5.3.3 Design of a Closed-Loop Neural Interface with a PID Controller 186 5.3.3.1 Introduction 186 5.3.3.2 System and Circuit Implementation 187 5.3.3.3 Experimental Results 189 6 System Integration and Experiments 193 6.1 Introduction 193 6.2 The PennBMBI: A General–Purpose Experimental Platform 194 6.2.1 Introduction 194 6.2.2 System Overview 195 6.2.3 Hardware Implementation 197 6.2.3.1 Neural Signal Analyzer 197 6.2.3.2 Neural Stimulators 200 6.2.3.3 Body Area Sensors 201 6.2.3.4 Computer Interface 202 6.2.4 Experimental Results 203 6.2.4.1 Bench Testing 203 6.2.4.2 In Vivo Experiments 206 6.3 The Watermaze 207 6.3.1 Introduction and Background 207 6.3.2 System Overview 209 6.3.3 Hardware Implementation 210 6.3.3.1 Design of the Watermaze Stimulator 210 6.3.3.2 Electrode and Electrode Connector 217 6.3.3.3 Image Sensor and Computer Interface 217 6.3.4 Software Implementation 218 6.3.4.1 Communication Protocol 218 6.3.4.2 Animal Tracking 221 6.3.4.3 User Interface 222 6.3.5 Experimental Results 224 6.4 Bidirectional Neural Interface for Freely Behaving Macaque 227 6.4.1 Introduction and Background 227 6.4.2 Circuit and System Design 228 6.4.3 Experimental Results 234 7 Conclusion and Future Direction 245 7.1 Summary of the Work 245 7.2 Future Direction 247 Bibliography 248 Index 263 Front Matter ....Pages i-xxxiii Introduction (Xilin Liu, Jan Van der Spiegel)....Pages 1-15 Neural Recording Front-End Design (Xilin Liu, Jan Van der Spiegel)....Pages 17-68 Neural Feature Extraction (Xilin Liu, Jan Van der Spiegel)....Pages 69-102 Neural Stimulator Design (Xilin Liu, Jan Van der Spiegel)....Pages 103-135 Bidirectional Neural Interface and Closed-Loop Control (Xilin Liu, Jan Van der Spiegel)....Pages 137-164 System Integration and Experiments (Xilin Liu, Jan Van der Spiegel)....Pages 165-216 Conclusion and Future Direction (Xilin Liu, Jan Van der Spiegel)....Pages 217-219 Back Matter ....Pages 221-242