Artificial Neural Networks in BiomedicinePaulo J.G. Lisboa, Emmanuel C. Ifeachor, Piotr S. Szczepaniak Following the intense research activIties of the last decade, artificial neural networks have emerged as one of the most promising new technologies for improving the quality of healthcare. Many successful applications of neural networks to biomedical problems have been reported which demonstrate, convincingly, the distinct benefits of neural networks, although many ofthese have only undergone a limited clinical evaluation. Healthcare providers and developers alike have discovered that medicine and healthcare are fertile areas for neural networks: the problems here require expertise and often involve non-trivial pattern recognition tasks - there are genuine difficulties with conventional methods, and data can be plentiful. The intense research activities in medical neural networks, and allied areas of artificial intelligence, have led to a substantial body of knowledge and the introduction of some neural systems into clinical practice. An aim of this book is to provide a coherent framework for some of the most experienced users and developers of medical neural networks in the world to share their knowledge and expertise with readers. |
Contents
The Bayesian Paradigm Second Generation Neural Computing | 11 |
12 Theory | 12 |
121 Bayesian Learning | 13 |
122 The Evidence Framework | 15 |
1221 Error bars | 16 |
1223 Regularisation | 17 |
123 Committees | 18 |
13 Example Results | 20 |
115 Conclusion | 149 |
Signal Processing | 151 |
Independent Components Analysis | 153 |
1221 The Decorrelating Manifold | 155 |
1222 The Choice of NonLinearity | 156 |
1223 ModelOrder Estimation | 158 |
1231 Illustration | 161 |
124 Applications | 163 |
14 Conclusion | 21 |
The Role of the Artificial Neural Network in the Characterisation of Complex Systems and the Prediction of Disease | 25 |
21 Introduction | 26 |
22 Diagnosis of Disease | 28 |
23 Outcome Prediction | 30 |
24 Conclusion | 31 |
Genetic Evolution of Neural Network Architectures | 39 |
The BiasVariance Problem | 40 |
33 Genetic Algorithms and Artificial Neural Networks | 41 |
331 Description of a General Method for Evolving ANN Architecture EANN | 42 |
332 Prediction of Depression After Mania | 43 |
334 ANN and the StabilitySpecialisation Choice | 45 |
34 Conclusion | 46 |
Computer Aided Diagnosis | 49 |
The Application of PAPNET to Diagnostic Cytology | 51 |
42 First Efforts at Automation in Cytology | 52 |
43 Neural Networks | 53 |
441 Components of the PAPNET System | 54 |
4411 Technical factors affecting the performance of the machine | 59 |
443 Application of the PAPNET System to Smears of Sputum | 61 |
445 Application of the PAPNET System to Oesophageal Smears | 62 |
45 Comment | 63 |
ProstAsure Index A SerumBased Neural NetworkDerived Composite Index for Early Detection of Prostate Cancer | 69 |
52 Clinical Background of Prostate Cancer and Derivation of the ProstAsure Index Algorithm | 70 |
53 Validation of PI with Independent Clinical Data | 72 |
54 Issues in Developing PI | 73 |
55 Conclusion | 76 |
Neurometric Assessment of Adequacy of Intraoperative Anaesthetic | 81 |
62 Measuring Sensory Perception | 82 |
64 Results | 83 |
65 Implementation | 86 |
66 Clinical Deployment | 88 |
67 Healthcare Benefit | 89 |
Classifying Spinal Measurements Using a Radial Basis Function Network | 93 |
72 Data | 94 |
723 Preprocessing the Data | 95 |
73 Radial Basis Functions and Networks | 96 |
74 Matrix Notation | 97 |
75 Training RBF Networks | 98 |
7522 Calculating the regularisation coefficients and the weights | 99 |
7523 Forward subset selection of RBFs | 101 |
7524 Input feature selection | 102 |
77 Conclusion | 103 |
GEORGIA An Overview | 105 |
81 Introduction | 106 |
82 The Medical Decision Support System | 107 |
83 Learning Pattern Generation | 109 |
84 Software and Hardware Implementation | 110 |
85 ReTraining and ReConfiguring the MDSS | 111 |
87 Conclusion | 114 |
Patient Monitoring Using an Artificial Neural Network | 117 |
92 Basic Statistical Appraisal of Vital Function Data | 118 |
93 Neural Network Details | 120 |
931 Default Training | 121 |
94 Implementation | 123 |
95 Clinical Trials | 124 |
96 Clinical Practice | 125 |
Benchmark of Approaches to Sequential Diagnosis | 129 |
102 Preliminaries | 130 |
103 Methods | 132 |
103 12 The diagnostic algorithm for second order markov chains the Markov II algorithm | 133 |
1032 The Fuzzy Methods | 135 |
10325 The reduced algorithm with secondorder context fuzzy 2B | 136 |
104 A Practical Example Comparative Analysis of Methods | 137 |
105 Conclusion | 138 |
Application of Neural Networks in the Diagnosis Of Pathological Speech | 141 |
112 The Research Material and the Problems Considered | 142 |
1122 Maxillofacial Surgery | 143 |
1123 Orthodontics | 144 |
113 The Signal Parameterisation | 145 |
114 The Application of the Neural Networks and the Results | 147 |
1241 Source Separation | 164 |
125 Conclusion | 166 |
Rest EEG Hidden Dynamics as a Discriminant for Brain Tumour Classification | 169 |
131 Introduction | 170 |
132 Characterising Hidden Dynamics | 171 |
133 The Clinical Study | 174 |
134 The Minimum Markov Order | 176 |
135 Conclusion | 179 |
Artificial Neural Network Control on Functional Electrical Stimulation Assisted Gait for Persons with Spinal Cord Injury | 181 |
141 Introduction | 182 |
142 Methods | 183 |
143 Results | 187 |
144 Discussion | 191 |
The Application of Neural Networks to Interpret Evoked Potential Waveforms | 195 |
152 The Medical Conditions Studied | 196 |
154 The Relationship Between the CNV and the Medical Conditions | 197 |
155 Experimental Procedures | 198 |
157 Feature Extraction | 199 |
158 Normalisation | 200 |
1592 The Probabilistic Simplified Fuzzy ARTMAP | 204 |
1593 ANN Training and Accuracy | 205 |
15933 Committees of ANNs | 206 |
1511 Results | 207 |
1513 Future Developments | 208 |
Image Processing | 211 |
Intelligent Decision Support Systems in the Cytodiagnosis of Breast Carcinoma | 213 |
162 Previous Work on Decision Support in this Domain | 215 |
1632 Input Variables | 216 |
1633 Partitioning of the Data | 217 |
165 Logistic Regression | 218 |
166 Data Derived Decision Tree | 219 |
167 MultiLayer Perceptron Neural Networks | 220 |
168 Adaptive Resonance Theory Mapping ARTMAP Neural Networks | 222 |
1683 Results from the Cascaded System | 225 |
169 Assessment of the Different Decision Support Systems | 226 |
A NeuralBased System for the Automatic Classification and FollowUp of Diabetic Retinopathies | 233 |
172 The DRA System | 235 |
173 Hybrid Module | 237 |
174 Committee Algorithms | 239 |
1741 New Selection Algorithms | 240 |
17411 Greedy selection | 241 |
1742 Sequential Cooperation | 242 |
1743 Experimental Results | 243 |
175 Related Work | 245 |
177 Conclusion | 246 |
Classification of Chromosomes A Comparative Study of Neural Network and Statistical Approaches | 249 |
1812 Chromosome Classification | 250 |
1813 Experimental Data | 251 |
182 The Neural Network Classifier | 252 |
1822 Network Topology and Training | 253 |
1823 Incorporating NonBanding Features | 254 |
183 Classification Performance | 255 |
1832 Comparison with Statistical Classifiers | 256 |
184 The Use of Context in Classification | 258 |
1842 Applying the Constraint by a Network | 259 |
1843 Results of Applying the Context Network | 260 |
185 Conclusion and Discussion | 261 |
1852 Training Set Size and Application of Context | 262 |
1853 Biological Context | 263 |
The Importance of Features and Primitives for MultidimensionalMultichannel Image Processing | 267 |
192 The Image Data Level | 269 |
194 Region Segmentation Quality and Training Phase | 270 |
195 Validation of Image Segmentation | 271 |
196 Segmentation Complexity and Quantitative Error Evaluation | 273 |
197 Feature Description | 275 |
198 Feature Selection | 276 |
199 A Preliminary Overview of Application Results | 278 |
1910 Conclusion | 280 |
| 281 | |
Other editions - View all
Artificial Neural Networks in Biomedicine Paulo J G Lisboa,Emmanuel C Ifeachor,Piotr S Szczepaniak No preview available - 2000 |
Common terms and phrases
abnormal accuracy alarm algorithm anaesthesia anaesthetic analysis application approach approximate architecture artificial neural network automated automatic Bayesian benign biomedical brain cells centromere Chapter chromosomes classification clinical committee complex components continuous function cytology Cytopathology data set decision boundary decision support density detection developed a network disease error estimation evaluation false feature selection feature space Figure Functional Electrical Stimulation Gaussian genetic algorithm heel switch hidden dynamics implementation Independent Components Analysis karyotyping laboratory layer learning lesion limit-alarms linear logistic regression malignant matrix measurements methods minimise monitor network to predict Neural Computation nodes non-linear nuclei optimum oscillations outcome output PAPNET System parameters pathological speech patients pattern recognition performance posterior posterior probability problem prostate cancer regularisation rules samples screening segmentation sensor sets sequential diagnosis SFAM signal smears specific spinal statistical structure techniques training set tumour Validation variables weights
Popular passages
Page vi - Fondaparinux has been approved by the US Food and Drug Administration (FDA) for the treatment of VTE.
References to this book
Neural Networks in Healthcare: Potential and Challenges Rezaul Begg,Joarder Kamruzzaman,Ruhul Sarkar No preview available - 2006 |