Computational Intelligence and Healthcare Informatics E-Book

• Herausgeber: John Wiley & Sons
• Kategorie: Wissenschaft und neue Technologien
• Serie: Machine Learning in Biomedical Science and Healthcare Informatics
• Sprache: Englisch

Seitenzahl: 706

Veröffentlichungsjahr: 2021

Beschreibung

COMPUTATIONAL INTELLIGENCE and HEALTHCARE INFORMATICS The book provides the state-of-the-art innovation, research, design, and implements methodological and algorithmic solutions to data processing problems, designing and analysing evolving trends in health informatics, intelligent disease prediction, and computer-aided diagnosis. Computational intelligence (CI) refers to the ability of computers to accomplish tasks that are normally completed by intelligent beings such as humans and animals. With the rapid advance of technology, artificial intelligence (AI) techniques are being effectively used in the fields of health to improve the efficiency of treatments, avoid the risk of false diagnoses, make therapeutic decisions, and predict the outcome in many clinical scenarios. Modern health treatments are faced with the challenge of acquiring, analyzing and applying the large amount of knowledge necessary to solve complex problems. Computational intelligence in healthcare mainly uses computer techniques to perform clinical diagnoses and suggest treatments. In the present scenario of computing, CI tools present adaptive mechanisms that permit the understanding of data in difficult and changing environments. The desired results of CI technologies profit medical fields by assembling patients with the same types of diseases or fitness problems so that healthcare facilities can provide effectual treatments. This book starts with the fundamentals of computer intelligence and the techniques and procedures associated with it. Contained in this book are state-of-the-art methods of computational intelligence and other allied techniques used in the healthcare system, as well as advances in different CI methods that will confront the problem of effective data analysis and storage faced by healthcare institutions. The objective of this book is to provide researchers with a platform encompassing state-of-the-art innovations; research and design; implementation of methodological and algorithmic solutions to data processing problems; and the design and analysis of evolving trends in health informatics, intelligent disease prediction and computer-aided diagnosis. Audience The book is of interest to artificial intelligence and biomedical scientists, researchers, engineers and students in various settings such as pharmaceutical & biotechnology companies, virtual assistants developing companies, medical imaging & diagnostics centers, wearable device designers, healthcare assistance robot manufacturers, precision medicine testers, hospital management, and researchers working in healthcare system.

Leseprobe

Table of Contents

Cover

Title Page

Copyright

Preface

Part I: Introduction

1 Machine Learning and Big Data: An Approach Toward Better Healthcare Services

1.1 Introduction

1.2 Machine Learning in Healthcare

1.3 Machine Learning Algorithms

1.4 Big Data in Healthcare

1.5 Application of Big Data in Healthcare

1.6 Challenges for Big Data

1.7 Conclusion

References

Part II: Medical Data Processing and Analysis

2 Thoracic Image Analysis Using Deep Learning

2.1 Introduction

2.2 Broad Overview of Research

2.3 Existing Models

2.4 Comparison of Existing Models

2.5 Summary

2.6 Conclusion and Future Scope

References

3 Feature Selection and Machine Learning Models for High-Dimensional Data: State-of-the-Art

3.1 Introduction

3.2 Types of Feature Selection

3.3 Machine Learning and Deep Learning Models

3.4 Real-World Applications and Scenario of Feature Selection

3.5 Conclusion

References

4 A Smart Web Application for Symptom-Based Disease Detection and Prediction Using State-of-the-Art ML and ANN Models

4.1 Introduction

4.2 Literature Review

4.3 Dataset, EDA, and Data Processing

4.4 Machine Learning Algorithms

4.5 Work Architecture

4.6 Conclusion

References

5 Classification of Heart Sound Signals Using Time-Frequency Image Texture Features

5.1 Introduction

5.2 Related Work

5.3 Theoretical Background

5.4 Proposed Algorithm

5.5 Experimental Results

5.6 Conclusion

References

6 Improving Multi-Label Classification in Prototype Selection Scenario

6.1 Introduction

6.2 Related Work

6.3 Methodology

6.4 Performance Evaluation

6.5 Experiment Data Set

6.6 Experiment Results

6.7 Conclusion

References

7 A Machine Learning–Based Intelligent Computational Framework for the Prediction of Diabetes Disease

7.1 Introduction

7.2 Materials and Methods

7.3 Machine Learning Classification Hypotheses

7.4 Classifier Validation Method

7.5 Performance Evaluation Metrics

7.6 Results and Discussion

7.7 Conclusion

References

8 Hyperparameter Tuning of Ensemble Classifiers Using Grid Search and Random Search for Prediction of Heart Disease

8.1 Introduction

8.2 Related Work

8.3 Proposed Method

8.4 Experimental Outcomes and Analyses

8.5 Conclusion

References

9 Computational Intelligence and Healthcare Informatics Part III—Recent Development and Advanced Methodologies

9.1 Introduction: Simulation in Healthcare

9.2 Need for a Healthcare Simulation Process

9.3 Types of Healthcare Simulations

9.4 AI in Healthcare Simulation

9.5 Conclusion

References

10 Wolfram’s Cellular Automata Model in Health Informatics

10.1 Introduction

10.2 Cellular Automata

10.3 Application of Cellular Automata in Health Science

10.4 Cellular Automata in Health Informatics

10.5 Health Informatics–Deep Learning–Cellular Automata

10.6 Conclusion

References

Part III: Machine Learning and COVID Prospective

11 COVID-19: Classification of Countries for Analysis and Prediction of Global Novel Corona Virus Infections Disease Using Data Mining Techniques

11.1 Introduction

11.2 Literature Review

11.3 Data Pre-Processing

11.4 Proposed Methodologies

11.5 Experimental Results

11.6 Conclusion and Future Scopes

References

12 Sentiment Analysis on Social Media for Emotional Prediction During COVID-19 Pandemic Using Efficient Machine Learning Approach

12.1 Introduction

12.2 Literature Review

12.3 System Design

12.4 Result and Discussion

12.5 Conclusion

References

13 Primary Healthcare Model for Remote Area Using Self-Organizing Map Network

13.1 Introduction

13.2 Background Details and Literature Review

13.3 Methodology

13.4 Results and Discussion

13.5 Conclusion

References

14 Face Mask Detection in Real-Time Video Stream Using Deep Learning

14.1 Introduction

14.2 Related Work

14.3 Proposed Work

14.4 Results and Evaluation

14.5 Conclusion

References

15 A Computational Intelligence Approach for Skin Disease Identification Using Machine/Deep Learning Algorithms

15.1 Introduction

15.2 Research Problem Statements

15.3 Dataset Description

15.4 Machine Learning Technique Used for Skin Disease Identification

15.5 Result and Analysis

15.6 Conclusion

References

16 Asymptotic Patients’ Healthcare Monitoring and Identification of Health Ailments in Post COVID-19 Scenario

16.1 Introduction

16.2 Material Properties and Design Specifications

16.3 Experimental Methods and Materials

16.4 Simulation Results

16.5 Conclusion

16.6 Abbreviations and Acronyms

References

17 COVID-19 Detection System Using Cellular Automata–Based Segmentation Techniques

17.1 Introduction

17.2 Literature Survey

17.3 Proposed Methodology

17.4 Results and Discussion

17.5 Conclusion

References

18 Interesting Patterns From COVID-19 Dataset Using Graph-Based Statistical Analysis for Preventive Measures

18.1 Introduction

18.2 Methods

18.3 GSA Model: Graph-Based Statistical Analysis

18.4 Graph-Based Analysis

18.5 Machine Learning Techniques

18.6 Exploratory Data Analysis

18.7 Conclusion

18.8 Limitations

Acknowledgments

Abbreviations

References

Part IV: Prospective of Computational Intelligence in Healthcare

19 Conceptualizing Tomorrow’s Healthcare Through Digitization

19.1 Introduction

19.2 Importance of IoMT in Healthcare

19.3 Case Study I: An Integrated Telemedicine Platform in Wake of the COVID-19 Crisis

19.4 Case Study II: A Smart Sleep Detection System to Track the Sleeping Pattern in Patients Suffering From Sleep Apnea

19.5 Future of Smart Healthcare

19.6 Conclusion

References

20 Domain Adaptation of Parts of Speech Annotators in Hindi Biomedical Corpus: An NLP Approach

20.1 Introduction

20.2 Review of Related Literature

20.3 Scope and Objectives

20.4 Methodological Design

20.5 Evaluation

20.6 Issues

20.7 Conclusion and Future Work

Acknowledgements

References

21 Application of Natural Language Processing in Healthcare

21.1 Introduction

21.2 Evolution of Natural Language Processing

21.3 Outline of NLP in Medical Management

21.4 Levels of Natural Language Processing in Healthcare

21.5 Opportunities and Challenges From a Clinical Perspective

21.6 Openings and Difficulties From a Natural Language Processing Point of View

21.7 Actionable Guidance and Directions for the Future

21.8 Conclusion

References

Index

End User License Agreement

List of Figures

Chapter 1

Figure 1.1 Machine learning and big data analysis in healthcare.

Figure 1.2 Application of ML in healthcare.

Figure 1.3 The types of machine learning algorithm.

Figure 1.4 Sources of big data in healthcare.

Figure 1.5 Applications of big data in healthcare.

Chapter 2

Figure 2.1 Broad view of existing research.

Figure 2.2 Types of chest pathologies.

Chapter 3

Figure 3.1 (a) This what the normal data looks like. (b) “Big p and small n” pro...

Figure 3.2 Feature selection process.

Figure 3.3 Taxonomy of feature selection.

Figure 3.4 Taxonomy of deep learning models.

Chapter 4

Figure 4.1 Overview of the processes involved in building ML models.

Figure 4.2 Machine Learning vs. Deep Learning workflow process.

Figure 4.3 Overview of an Artificial Neural Network.

Figure 4.4 The description of the raw dataset showing the first five rows.

Figure 4.5 The code snippet showing the data processing of the attributes such a...

Figure 4.6 The description of the cleaned and processed dataset showing the firs...

Figure 4.7 Boxplot showing number of disease occurrences.

Figure 4.8 Ten highest reported diseases.

Figure 4.9 Ten highest reported symptoms.

Figure 4.10 The code snippet of the label encoding and one hot encoding process ...

Figure 4.11 List of the implemented algorithms while building the model. MLP pro...

Figure 4.12 Work architecture of the proposed solution.

Figure 4.13 The web application to predict disease based on symptoms.

Figure 4.14 The user can input multiple symptoms at a time and get accurate pred...

Chapter 5

Figure 5.1 PCG signal in time-frequency representation before and after filtered...

Figure 5.2 (a) Standard spectrogram for normal sample. (b) for standard spectrog...

Figure 5.3 (a) Mel-spectrogram for normal sample. (b) Mel-spectrogram for abnorm...

Figure 5.4 (a) IIR-CQT spectrogram for normal sample. (b) IIR-CQT spectrogram fo...

Figure 5.5 Kirschmask in eight different directions [12].

Figure 5.6 Algorithm for the proposed heart sound classification system.

Figure 5.7 Confusion matrix for dense CLBP method.

Chapter 6

Figure 6.1 Process of proposed model.

Figure 6.2 Performance on the Yeast data set.

Figure 6.3 Performance on the Scene data set.

Figure 6.4 Performance of the Emotion data set.

Figure 6.5 Performance if the Enron data set.

Figure 6.6 Performance if the Medical data set.

Chapter 7

Figure 7.1 An Intelligent Computational Framework for diabetes disease predictio...

Figure 7.2 Performance of all eight classification hypotheses on PIDD using 5-fo...

Figure 7.3 F1-score and MCC of all eight classification hypotheses on PIDD apply...

Figure 7.4 Precision and AUC of all eight classification hypotheses on PIDD usin...

Figure 7.5 ROC curves of all used classification hypotheses applying 5-fold CV.

Figure 7.6 Performance of all eight classification hypotheses on PIDD using 7-fo...

Figure 7.7 F1-score and MCC of all eight classification hypotheses on PIDD using...

Figure 7.8 Precision and AUC of all eight classification hypotheses on PIDD usin...

Figure 7.9 ROC curves of all used classification hypotheses using 7-fold CV.

Figure 7.10 Performance of all eight classification hypotheses on PIDD using 10-...

Figure 7.11 F1-score and MCC of all eight classification hypotheses on PIDD usin...

Figure 7.12 Precision and AUC of all eight classification hypotheses on PIDD usi...

Figure 7.13 ROC curves of all eight classification hypotheses on PIDD using 10-f...

Chapter 8

Figure 8.1 General schematic diagram of the proposed method for predicting heart...

Figure 8.2 Process of building heart disease prediction model.

Figure 8.3 Proposed classifier workflow.

Figure 8.4 Random forest classifier workflow.

Figure 8.5 Comparison of three ensemble classifier accuracy before and after gri...

Chapter 9

Figure 9.1 Distribution of post-surgical complications in the selected dataset.

Figure 9.2 High-level ML model architecture.

Figure 9.3 ROC-AUC curve of post-surgical complication outcomes for SGD classifi...

Figure 9.4 ROC-AUC curve of post-surgical complication outcomes for SVM nu-SVC c...

Figure 9.5 ROC-AUC Curve of post-surgical complication outcomes for RF classifie...

Figure 9.6 Basic LSTM unit.

Figure 9.7 Bidirectional LSTM model network architecture.

Figure 9.8 Precision of surgical participant classes.

Chapter 10

Figure 10.1 Block diagram of n-cell periodic boundary cellular automata with hyb...

Figure 10.2 E-healthcare steps toward medical diagnosis or disease detection.

Figure 10.3 A typical system architecture based on cellular automata used in hea...

Figure 10.4 Basic block diagram of Cellular Automata for Symbolic Induction (CAS...

Figure 10.5 Detection of heart disease with hybrid random forest with linear mod...

Figure 10.6 Basic block diagram of HL-DL-CA used in health informatics.

Chapter 11

Figure 11.1 Performance graph of LMBP algorithm Mean Square Error (MSE) vs. epoc...

Figure 11.2 Rise-wise classification of countries.

Chapter 12

Figure 12.1 Architecture diagram of COVID-19 emotional classification.

Figure 12.2 Pseudocode for Independent Component Analysis pre-processing (Prasty...

Figure 12.3 Pseudocode for WMAR features extraction (Satu, Md Shahriare,

et al

. ...

Figure 12.4 Chicken swarm optimization for COVID-19 emotional features optimizat...

Figure 12.5 Pseudocode for chicken swarm optimization (Yang, Liping, Alan M. Mac...

Figure 12.6 Structure of SVM.

Figure 12.7 Pair-wise comparison format.

Figure 12.8 Relative scale to compare COVID-19 two tweets.

Figure 12.9 Choosing COVID-19 emotional prediction.

Figure 12.10 Comparisons of accuracy, sensitivity, and specificity for various m...

Figure 12.11 Comparisons of precision, recall, and F-measure for various SA meth...

Figure 12.12 Comparison of processing time vs. no of products.

Chapter 13

Figure 13.1 Different steps in the block diagram of the proposed methodology.

Figure 13.2 Kohonen SOM network model.

Figure 13.3 Membership of features.

Figure 13.4 ROC curve of proposed methodology.

Figure 13.5 (a) SOM Hit map. (b) SOM weight for input vectors. (c) SOM weight po...

Chapter 14

Figure 14.1 Block diagram for proposed work.

Figure 14.2 Dataset distribution in training set.

Figure 14.3 Dataset distribution in testing set.

Figure 14.4 Proposed VGG19.

Figure 14.5 VGG19 layered architecture.

Figure 14.6 Experiment history for each epoch.

Figure 14.7 Accuracy curve.

Figure 14.8 Loss curve.

Figure 14.9 Normalized confusion matrix.

Figure 14.10 Confusion matrix without normalization.

Figure 14.11 Model prediction for sample images extracted from video stream.

Chapter 15

Figure 15.1 Actinic keratosis.

Figure 15.2 Melanoma.

Figure 15.3 Pigmented keratosis.

Figure 15.4 Nevus.

Figure 15.5 Vascular lesion.

Figure 15.6 SVM block diagram.

Figure 15.7 Recurrent neural network architecture.

Figure 15.8 Decision tree architecture.

Figure 15.9 CNN architecture.

Figure 15.10 Random Forest.

Figure 15.11 Random Forest architecture.

Figure 15.12 Accuracy rate.

Chapter 16

Figure 16.1 Block diagram.

Figure 16.2 (a) LM-35 temperature sensor, (b) AD8232 heart rate sensor, (c) puls...

Figure 16.3 Circuit diagram.

Figure 16.4 Data flow diagram.

Figure 16.5 A working prototype.

Figure 16.6 Reading on website.

Figure 16.7 Heart rate graph.

Figure 16.8 Pulse rate graph.

Figure 16.9 Temperature graph.

Figure 16.10 Gyroscopic inclination graph.

Figure 16.11 Database entries: value1, temperature; value2, heart rate; value3, ...

Chapter 17

Figure 17.1 Structure of 1D CA.

Figure 17.2 Structure of 2D CA [7].

Figure 17.3 (a) Von Neumann neighborhood models, (b) Moore neighborhood models, ...

Figure 17.4 CNN architecture.

Figure 17.5 System flow diagram.

Figure 17.6 Schema of convolutional neural network part.

Figure 17.7 Schema of artificial neural network part.

Figure 17.8 Normal chest X-ray [20].

Figure 17.9 COVID-19 positive chest X-ray [19].

Figure 17.10 Pneumonia chest X-ray [20].

Figure 17.11 Sample images of hidden layer 1.

Chapter 18

Figure 18.1 Data collection and cleaning mechanism.

Figure 18.2 Proposed GSA model.

Figure 18.3 Graph representation of healthcare data.

Figure 18.4 RDF data representation.

Figure 18.5 Knowledge representation of COVID, KaTrace dataset.

Figure 18.6 RDF for KaTrace, initial data only.

Figure 18.7 District-wise patient represented by age.

Figure 18.8 Betweenness centrality for knowledge graph.

Figure 18.9 PageRank centrality for knowledge graph.

Figure 18.10 RDF query on the knowledge graph.

Figure 18.11 PageRank centrality for knowledge graph.

Figure 18.12 Overall inter-district graph.

Figure 18.13 Reference graph for P566.

Figure 18.14 P653 spread to 45 nodes.

Figure 18.15 P653 infect details.

Figure 18.16 P4184, six-level parent-child relationships, and P4184 nodes connec...

Figure 18.17 Child nodes of Congregation and parent nodes of Congregation patien...

Figure 18.18 Part of the decision tree based on the apriori algorithm.

Figure 18.19 Association rule of data.

Figure 18.20 Model comparison.

Figure 18.21 ROC for decision tree classifier.

Figure 18.22 Age-wise case distribution.

Figure 18.23 Cases from Maharashtra.

Figure 18.24 Indicating cluster and reason attribute.

Figure 18.25 Primary contact tracing.

Figure 18.26 New cases forecast and active cases forecast.

Figure 18.27 New cases trend and weekly curve and active cases trend and weekly ...

Figure 18.28 New, active, and sample test curve.

Figure 18.29 Distribution of age.

Figure 18.30 Distribution of days.

Figure 18.31 PDF and CDF.

Figure 18.32 Box plot for survival rate.

Figure 18.33 Cases for 1 lakh population and ratio curve for total test and conf...

Figure 18.34 Ratio curve of sample tests over new cases and ratio curve predicti...

Figure 18.35 Parent-child spread curve and cases represented age-wise.

Figure 18.36 Day-wise cure status.

Figure 18.37 Gender-wise infection.

Figure 18.38 Curve rate over at a time.

Figure 18.39 Bangalore urban age-wise spread.

Figure 18.40 Cases for one lakh population and primary contact spread of major d...

Figure 18.41 Oversea patient’s district-wise count and patient from Maharashtra ...

Figure 18.42 Parent-child of the patient from Maharashtra.

Figure 18.43 Parent-child with a reason.

Figure 18.44 Death trend with reason.

Chapter 19

Figure 19.1 Our three-strata approach for building a strategic telemedicine plat...

Figure 19.2 Depiction of the flowchart of the “Homecare” model for our suggested...

Figure 19.3 Modus operandi post risk evaluation based on the Homecare model for ...

Figure 19.4 Integration of important components of the information flow that hea...

Figure 19.5 Workflow for the community model for the proposed design.

Figure 19.6 Depiction of the objectives of a possible surveillance system of the...

Figure 19.7 Illustration of a possible ‘contact tracing’ strategy for suspected ...

Figure 19.8 Classification of sleep apnea according to the different causes of o...

Figure 19.9 Illustration depicting how a person suffering from OSA different tha...

Figure 19.10 Depiction of the proposed design to effectively work in detection o...

Chapter 20

Figure 20.1 Systematic process of annotation (adapted from Behera, 2017).

Figure 20.2 ILCI ANN App v. 2.0.

Figure 20.3 Feature template of the SVM POS tagger model.

Figure 20.4 Typological statistics of errors.

Figure 20.5 A comparison of error rates between HMT and ABDT.

Chapter 21

Figure 21.1 Layout of Natural Language Processing.

Figure 21.2 Chronology of Natural Language Processing.

Figure 21.3 Abbreviation and its expansion used in the above article.

Figure 21.4 Levels of Natural Language Processing in healthcare.

Figure 21.5 Use of Natural Language Processing in clinical research.

Figure 21.6 Solution to the challenges in Natural Language Processing.

List of Table

Chapter 1

Table 2.1 Details of ChestX-ray14 dataset.

Table 2.2 Comparison of different deep learning models.

Table 2.3 Comparison of models on the basis of AUC score for 14 chest pathologie...

Table 2.4 Comparison of DL models on the basis of different performance metrics.

Table 2.5 Models with hardware used and time required for training.

Chapter 5

Table 5.1 Dataset description collected from PhysioNet 2016 challenge.

Table 5.2 Accuracies for conventional spectrogram generation based on various fe...

Table 5.3 Accuracies for various spectrogram generation techniques for WLD featu...

Table 5.4 Accuracies for different types of feature extraction method by Mel-spe...

Table 5.5 Comparison of proposed work with the related work.

Chapter 6

Table 6.1 Experimental data sets description.

Table 6.2 Comparative study of Hamming Loss of proposed (at 20% reduction) with ...

Table 6.3 Comparative study of One-Error of proposed (20% reduction) with ML-KNN...

Table 6.4 Comparative study of average precision of proposed (20% reduction) wit...

Table 6.5 Experimental result 1: Hamming Loss ↓.

Table 6.6 Experimental result 2: One-Error ↓.

Table 6.7 Experimental result 3: Average Precision ↑.

Chapter 7

Table 7.1 Information regarding Pima Indian diabetes dataset.

Table 7.2 Confusion matrix.

Table 7.3 Outcomes of all eight classification hypotheses on PIDD using 5-fold C...

Table 7.4 Performance of all eight classification hypotheses on PIDD using 7-fol...

Table 7.5 Performance of all eight classification hypotheses on PIDD using 10-fo...

Chapter 8

Table 8.1 Grid search hyperparameter tuning of ensemble classifier with hyperpar...

Table 8.2 Random search hyperparameter tuning of ensemble classifier with hyperp...

Table 8.3 UCI heart disease dataset properties.

Table 8.4 Performance evaluation measure of the proposed model.

Table 8.5 The outcomes of random forest (RF) before and after applying grid sear...

Table 8.6 The performance of AdaBoost (AB) classifier before and after applying ...

Table 8.7 The performance of gradient boosting (GB) classifier before and after ...

Chapter 9

Table 9.1 Healthcare simulation instances.

Table 9.2 Clinical characteristics of surgical patients treated for hematologica...

Table 9.3 Misclassification—Confusion matrix for bi-LSTM model.

Table 9.4 Misclassification—Confusion matrix for LR model.

Table 9.5 Misclassification—Confusion matrix for RF model.

Chapter 10

Table 10.1 Truth table.

Table 10.2 Application of cellular automata in health sciences.

Table 10.3 Cellular automata modeling in health sciences.

Table 10.4 Deep learning applications in health informatics.

Chapter 11

Table 11.1 COVID-19 dataset attributes in an abbreviated form [15, 21].

Table 11.2 Sample of range-wise classification of attributes [15, 21].

Table 11.3 Classification of correlation values [11].

Table 11.4 Correlation between variables (predictor vs. response).

Table 11.5 Straight line trend with regression coefficients.

Table 11.6 Sample of frequency count of variables.

Table 11.7 Maximum number of rule generation based on support values 20%, 30%, 4...

Table 11.8 Some interesting rules for COVID-19 dataset.

Table 11.9 Performance assessment of the proposed model.

Table 11.10 Performance of the proposed model.

Table 11.11 Country-wise performance comparison of actual and predicted output b...

Table 11.12 Country-wise performance comparison of actual and predicted output b...

Table 11.13 Risk classification of output parameters.

Table 11.14 Sample of risk-wise categorization of countries based on output para...

Chapter 12

Table 12.1 Scale for comparing alternatives.

Table 12.2 Number of comparisons.

Table 12.3 Comparisons of precision, recall, and F-measure for various sentiment...

Table 12.4 Comparisons of precision, recall, and F-measure for various sentiment...

Table 12.5 Time consumption of various sentiment analysis schemes (Satu, Md Shah...

Chapter 13

Table 13.1 Discretized features represented using fuzzy variables.

Table 13.2 Different severity condition.

Table 13.3 Calculation of severity factor.

Table 13.4 Performance measure with patients samples.

Chapter 14

Table 14.1 Classification report (in percent).

Chapter 18

Table 18.1 COVID-19 dataset “KaTrace” description.

Table 18.2 Relationship established in neo2j tool.

Table 18.3 Centrality for the knowledge graph.

Table 18.4 District-wise parent and child count of patients attended congregatio...

Table 18.5 Comparing the death count of a few districts.

Table 18.6 Survival status and gender count.

Table 18.7 Statistical inference.

Table 18.8 Data values of tests conducted.

Table 18.9 Parent and child relationship count.

Table 18.10 Spread count of a few patients.

Table 18.11 Oversea patient’s details.

Chapter 20

Table 20.1 IA agreement.

Table 20.2 Distribution of corpus in HMT and ABDT.

Table 20.3 Category-wise evaluation results of HMT and ABDT.

Guide

Cover

Table of Contents

Title Page

Copyright

Preface

Begin Reading

Index

End User License Agreement

Pages

v

ii

iii

iv

xv

xvi

xvii

1

3

4

5

6

7

8

9

10

11

12

13

15

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57

58

59

60

61

62

63

65

66

67

68

69

70

71

72

73

74

75

76

77

78

79

80

81

82

83

84

85

86

87

88

89

90

91

92

93

94

95

96

97

98

99

100

101

103

104

105

106

107

108

109

110

111

112

113

114

115

116

117

118

119

121

122

123

124

125

126

127

128

129

130

131

132

133

134

135

136

137

138

139

140

141

142

143

144

145

146

147

148

149

150

151

152

153

154

155

156

157

158

159

160

161

162

163

164

165

166

167

168

169

170

171

172

173

174

175

176

177

179

180

181

182

183

184

185

186

187

188

189

190

191

192

193

195

196

197

198

199

200

201

202

203

204

205

206

207

208

209

210

211

212

213

215

216

217

218

219

220

221

222

223

224

225

226

227

228

229

230

231

232

233

235

236

237

238

239

240

241

242

243

244

245

246

247

248

249

250

251

252

253

254

255

256

257

258

259

260

261

262

263

264

265

266

267

268

269

270

271

272

273

274

275

276

277

278

279

280

281

282

283

284

285

286

287

288

289

290

291

292

293

294

295

297

298

299

300

301

302

303

304

305

306

307

308

309

310

311

312

313

314

315

316

317

318

319

320

321

322

323

325

326

327

328

329

330

331

332

333

334

335

336

337

338

339

340

341

342

343

344

345

346

347

348

349

350

351

352

353

354

355

356

357

358

359

361

362

363

364

365

366

367

368

369

370

371

372

373

374

375

376

377

378

379

380

381

382

383

384

385

386

387

388

389

390

391

393

394

395

396

397

398

399

400

401

402

403

404

405

406

407

409

410

411

412

413

414

415

Scrivener Publishing

100 Cummings Center, Suite 541J

Beverly MA, 01915-6106

Machine Learning in Biomedical Science and Healthcare Informatics

Series Editors: Vishal Jain and Jyotir Moy Chatterjee

In this series, the focus centres on the various applications of machine learning in the biomedical engineering and healthcare fields, with a special emphasis on the most representative learning techniques, namely deep learning-based approaches. Machine learning tasks typically classified into two broad categories depending on whether there is a learning “label” or “feedback” available to a learning system: supervised learning and unsupervised learning. This series also introduces various types of machine learning tasks in the biomedical engineering field from classification (supervised learning) to clustering (unsupervised learning). The objective of the series is to compile all aspects of biomedical science and healthcare informatics, from fundamental principles to current advanced concepts.

Submission to the series: Please send book proposals to [email protected] and/or [email protected]

Publishers at Scrivener

Martin Scrivener ([email protected])

Phillip Carmical ([email protected])

Computational Intelligence and Healthcare Informatics

Edited by

Om Prakash Jena

Alok Ranjan Tripathy

Ahmed A. Elngar

Zdzislaw Polkowski

This edition first published 2021 by John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, USA and Scrivener Publishing LLC, 100 Cummings Center, Suite 541J, Beverly, MA 01915, USA

© 2021 Scrivener Publishing LLC

For more information about Scrivener publications please visit www.scrivenerpublishing.com.

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, except as permitted by law. Advice on how to obtain permission to reuse material from this title is available at http://www.wiley.com/go/permissions.

Wiley Global Headquarters111 River Street, Hoboken, NJ 07030, USA

For details of our global editorial offices, customer services, and more information about Wiley products visit us at www.wiley.com.

Limit of Liability/Disclaimer of WarrantyWhile the publisher and authors have used their best efforts in preparing this work, they make no representations or warranties with respect to the accuracy or completeness of the contents of this work and specifically disclaim all warranties, including without limitation any implied warranties of merchant-ability or fitness for a particular purpose. No warranty may be created or extended by sales representatives, written sales materials, or promotional statements for this work. The fact that an organization, website, or product is referred to in this work as a citation and/or potential source of further information does not mean that the publisher and authors endorse the information or services the organization, website, or product may provide or recommendations it may make. This work is sold with the understanding that the publisher is not engaged in rendering professional services. The advice and strategies contained herein may not be suitable for your situation. You should consult with a specialist where appropriate. Neither the publisher nor authors shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages. Further, readers should be aware that websites listed in this work may have changed or disappeared between when this work was written and when it is read.

Library of Congress Cataloging-in-Publication Data

ISBN 978-1-119-81868-7

Cover image: Pixabay.ComCover design by Russell Richadson

Set in size of 11pt and Minion Pro by Manila Typesetting Company, Makati, Philippines

Printed in the USA

10 9 8 7 6 5 4 3 2 1

Preface

Computational intelligence (CI) refers to the ability of computers to accomplish tasks that are normally completed by intelligent beings such as humans and animals. With the rapid advance of technology, artificial intelligence (AI) techniques are being effectively used in the fields of health to improve the efficiency of treatments, avoid the risk of false diagnoses, make therapeutic decisions, and predict the outcome in many clinical scenarios. Modern health treatments are faced with the challenge of acquiring, analysing and applying the large amount of knowledge necessary to solve complex problems. Computational intelligence in healthcare mainly uses computer techniques to perform clinical diagnoses and suggest treatments. In the present scenario of computing, CI tools present adaptive mechanisms that permit the understanding of data in difficult and changing environments. The desired results of CI technologies profit medical fields by assembling patients with the same types of diseases or fitness problems so that healthcare facilities can provide effectual treatments.

This book starts with the fundamentals of computer intelligence and the techniques and procedures associated with it. Contained in this book are state-of-the-art methods of computational intelligence and other allied techniques used in the healthcare system, as well as advances in different CI methods that will confront the problem of effective data analysis and storage faced by healthcare institutions. The objective of this book is to provide researchers with a platform encompassing state-of-the-art innovations; research and design; implementation of methodological and algorithmic solutions to data processing problems; and the design and analysis of evolving trends in health informatics, intelligent disease prediction and computer-aided diagnosis.

The book aims to integrate several aspects of CI, like machine learning and deep learning, from diversified perspectives involving recent research trends and advanced topics in the field which will be of interest to academicians and researchers working in this area. The purpose of the book is to endow different communities with innovative advances in theory, analytical approaches, numerical simulation, statistical analysis, modeling, advanced deployment, case studies, analytical results, computational structuring and significance progress in the field of machine learning and deep learning in healthcare applications. This book is targeted towards scientists, application doctors, health professionals, professors, researchers and students. Different dimensions of CI applications will be revealed and its use as a solution for assorted real-world biomedical and healthcare problems is illustrated. Following is a brief description of the subjects covered in each chapter.

–

Chapter 1

is a systematic review of better options in the field of healthcare using machine learning and big data. The use of machine learning (ML) and big data in several application areas in healthcare services are identified which can further improve the unresolved challenges. Technologies such as ML will greatly transform traditional healthcare services and improve the relationship between service users and providers, providing better service in less time. Moreover, ML will help in keeping an eye on critical patients in real time, diagnose their disease, and recommend further treatment.

–

Chapter 2

provides a critical analysis of the deep learning techniques utilized for thoracic image analysis and the respective accuracy achieved by it. Various deep learning techniques are described along with dataset, activation function and model used, number and types of layers used, learning rate, training time, epoch, performance metric, hardware used and type of abnormality detected. Moreover, a comparative analysis of existing deep learning models based on accuracy, precision and recall is also presented with an emphasis on the future direction of research.

–

Chapter 3

discusses various application domains in need of feature selection techniques and also the way to deal with feature reduction problems occurring in large, voluminous datasets.

–

Chapter 4

presents a detailed analysis of the available ML and ANN models conducted vis-à-vis the data considered, and the best one is applied for training and testing the neural networks developed for the present work to detect and predict a disease based on the symptoms described.

–

Chapter 5

presents an approach for heart sound classification using the time-frequency image texture feature and support vector machine classifier.

–

Chapter 6

proposes a novel approach for selecting a prototype without dropout for the accuracy of the multi-label classification algorithm.

–

Chapter 7

introduces an intelligent computational predictive system for the identification and diagnosis of diabetes. Here, eight machine learning classification hypotheses are examined for the identification and diagnosis of diabetes. Numerous performance measuring metrics, such as accuracy, sensitivity, specificity, AUC, F1-score, MCC, and ROC curve, are applied to inspect the effectiveness and stability of the proposed model.

–

Chapter 8

proposes hyperparameter optimization for ensemble learners as it has a lot of hyperparameters. The optimized ensemble learning model can be built by tuning the hyperparameters of ensemble learners. This chapter applies a grid search and random search algorithms for tuning the hyper-parameters of ensemble learners. Three ensemble learners are used in this proposed work: two boosting models (AdaBoost and Gradient boosting algorithms) and one bagging model (Random Forest algorithm).

–

Chapter 9

presents a detailed analysis of the different types of healthcare simulations—from discrete event simulation (DES) and agent-based methods (ABM) to system dynamics (SD).

–

Chapter 10

focuses on the application of Wolfram’s cellular automata (CA) model in different domains of health informatics, medical informatics and bioinformatics. It also reports on the analysis of medical imaging for breast cancer, heart disease, tumor detection and other diseases using CA. Augmenting the machine learning mechanism with CA is also discussed, which provides higher accuracy, precision, security, speed, etc.

–

Chapter 11

considers the global dataset of 204 countries for the period of December 31

st

2019 to May 19

th

2020 from the Worldometer website for study purpose and data from May 20

th

to June 8

th

is considered to predict the evaluation of the outbreak, i.e., three weeks ahead. Three of the most prominent data mining techniques—linear regression (LR), association rule mining (ARM) and back propagation neural network (BPNN)—are utilized to predict and analyze the COVID-19 dataset.

–

Chapter 12

proposes a hybrid support vector machine (SVM) with chicken swarm optimization (CSO) algorithm for efficient sentiment analysis. Part-of-speech (POS) tagged text is used in this algorithm for extracting the potential features.

–

Chapter 13

discusses the primary healthcare model for remote areas using a self-organizing map network.

–

Chapter 14

proposes a real-time face mask detection approach using VGG19 from the video stream recorded using a webcam that achieved 100 percent training accuracy with logloss 0.00 and a validation accuracy of 99.63 percent with logloss 0.01 in just 20 epochs.

–

Chapter 15

focuses on different types of machine and deep learning algorithms like CNN and SVM for skin disease classification. The methods are very helpful in identifying skin diseases very easily and in fewer time periods.

–

Chapter 16

discusses a program developed for collecting heart rhythm, pulse rate, body temperature, and inclination data from patients.

–

Chapter 17

describes a proposed automatic COVID-19 detection system that can be used as an alternative diagnostic medium for the virus.

–

Chapter 18

presents an innovative approach for deriving interesting patterns using machine learning and graph database models to incorporate the preventive measures in an earlier state. A graph-based statistical analysis (GSA) model to study the COVID-19 pandemic outbreak’s impact is also proposed.

–

Chapter 19

discusses the conceptualization of tomorrow’s healthcare through digitization. The objective of this chapter is to utilize the latest resources available at hand to design the case studies.

–

Chapter 20

provides a systematic procedure for the development of the POS tagger trained on general domain corpus and the development of biomedical corpus in Hindi.

–

Chapter 21

studies the concepts of neuro-linguistic programming (NLP) used in healthcare applications, and also examines the NLP system and the resources which are used in healthcare. Along with the challenges of NLP, different aspects of NLP are studied as well as clinical methods.

Finally, we would like to sincerely thank all the chapter authors who contributed their time and expertise and for helping to reach a successful completion of the book. Second, the editors wish to acknowledge the valuable contributions of the reviewers regarding the improvement of quality, coherence, and content presentation of chapters.

The EditorsJuly 2021

Part IINTRODUCTION

1Machine Learning and Big Data: An Approach Toward Better Healthcare Services

Nahid Sami* and Asfia Aziz

School of Engineering Science and Technology, Jamia Hamdard, New Delhi, India

Abstract

Artificial Intelligence has been considered the biggest technology in transforming society. The transformation is highly influenced by the tools and techniques provided by Machine Learning and Deep Learning. One latest discovery is Robotic Surgical Tools like Da Vinci robot which helps surgeons to perform surgeries more accurately and detect distances of particular body parts to perform surgery precisely over them using computer vision assisted by machine learning. The maintenance of health records is a tedious process and ML in association with Big Data has greatly saved the effort, money and time required while processing such records. MIT is working on intelligent smart health record learning method using machine learning techniques to give suggestions regarding diagnosis and medication. Deep mining techniques are used in medical image analysis for the detection of tissue in radiotherapy. Outbreak of several chronic diseases can also be predicted using data gathered through satellite, online information and social media updates. Big Data provides a major part by maintaining the EHR which is in form of complex unstructured data. Another major challenging area is related to the infrastructure of the hospital. Moving toward more advance technologies, the infrastructure need to be updated which is time consuming and costly.

Keywords: Machine learning, deep learning, healthcare, electronic medical records (EMRs), big data

1.1 Introduction

Machine Learning (ML) is a computer program that learns from experience with respect to a particular task assigned and gives result accordingly. The performance of such computational algorithm improves with experience. Health is a major area of concern for everyone and to provide the best healthcare service is becoming one of the major goals of almost every country. But doing that is not an easy task as collecting the medical data and providing it to leverage knowledge so that the best possible treatment can be provided is itself very challenging. So, data plays a crucial part in extracting information and addressing problems related to health. ML has the ability to extract information from the data being provided and further helps in resolving this fundamental issue to some extent.

The huge medical data need to be interpreted and processed by epidemiologists. The input of healthcare providers has been expanded and also created new opportunities due to the availability of huge amount of data related to patients and facility being provided which will further help in achieving the necessary approaches related to prevention and treatment [1]. Due to the complexity of medical data and also lack of technology, the collection was completely ignored in the past. ML algorithm has proved to overcome such difficulties by collecting the medical data securely and further applying it for diagnosis and prognosis. ML has improved several domains like Automatic Speech Recognition (ASR), Natural Language Processing (NLP), and computer vision by using the data. Creating the correct model for maintaining the electronic medical records (EMRs) is a challenging issue due to its availability, quality, and heterogeneity.

Big data is going to play a major role in revolutionizing the healthcare services in the coming future by using algorithm to detect and treat diseases [2]. Its impact on the practice of medicine is fundamentally going to transform the physician ability to personalize care directly to the people. The way to achieve this goal is by collecting data through handheld and wearable devices. This data will be compared with the genetic profile of people and further used for decision-making. The vast medical data needs to be integrated and accessed intelligently to support better healthcare delivery. Big data can create new networks of sharing knowledge by measuring and monitoring processes digitally [3]. Data comparison will be easier which will facilitate streamlined workflows greater efficiencies and improved patient care. Systematic analysis of extensive data can help to detect patterns so that clinicians can provide treatment to individuals and project health outcomes. Digital networks can bring together partners and knowledge sharing delivering context relevant clinical information enables more holistic decision-making. Healthcare can only benefit from big data when it is made structured relevant smart and accessible.

Figure 1.1 shows the how ML and big data analytics plays an important role in different fields associated with healthcare services. There are five major modules associated with ML algorithm and their contribution. The physician unstructured data is provided to ML algorithm and, in return, gets better clinical decision support. Also, the radiologist provides data in form of MRI/images and receives diagnostics from ML. It provides the patient a better lifestyle advice and treatment option. Patients are complex module with different genetic back ground so the risks associated with them are different over time. The drug makers get patients medical records for development of necessary drugs. The clinical research and development module provides bio illustration to the algorithm and gets predictive analysis.

1.2 Machine Learning in Healthcare

In recent years, artificial intelligence (AI) has shown tremendous growth in transforming every aspect of life due to its wide range of tools which help in decision-making by analyzing data and integrating information. In terms of technology, Al has stolen spotlight and its advancements are quicker than our prediction [4]. ML being a subset of AI is transforming the world and raising its importance for the society. ML is defined as the study of methods and tools which help in identifying patterns within data and make computer learn without being programmed explicitly. ML can further be used to extend our knowledge regarding current scenario as well as for future prediction by allowing program to learn through experience. It uses the concept of AI for data optimization. Analyzing the best model to make the machine intelligent for data explanation is the goal. We will be discussing here its development in the field of medicine.

Figure 1.1 Machine learning and big data analysis in healthcare.

Figure 1.2 shows the different areas where ML algorithm is playing a major role to provide better healthcare services. Applying such technology will help in proving personalized treatment which will improve the health condition of patients. Drug discovery and research will be highly benefited as the structured data will be available. Further support will be provided to clinical decision-making and early detection of diseases will be possible to make the services better for individual. The use of Deep Learning (DL) and neural network will be highly helpful in improving the imaging and diagnostic techniques. By proving all the essential services in medical, the fraud related to medical insurance will minimized to the least.

ML can transform the healthcare services by making us better providers of correct medical facility at the patient level. We can gather information on how different environmental exposure and lifestyle will vary the symptoms of disease. The intervention and history will help us decide treatment and decision-making. We can further understand the health and disease trajectory which will help in prepare us before arrival of the pandemics in worst possible situation. The resources available to us can be utilized in more efficient way with reduced costs. Also, the public health policies can be transformed in a way benefiting the people.

Figure 1.2 Application of ML in healthcare.

1.3 Machine Learning Algorithms

Depending upon the problem and approach to be applied, it has been categorized into various types among which major application lie into supervised, unsupervised, semi-supervised, reinforcement, and DL. The various types and its contribution to healthcare sector are shown in Figure 1.3.

1.3.1 Supervised Learning

This ML algorithm works under supervision, i.e., machine is trained with data which is well labeled and helps the model to predict with the help of dataset. Furthermore, supervised learning is divided into classification and regression. When the resultant variable is categorical, i.e., with two or more classes (yes/no, true/false, disease/no disease), we make use of classification. Whereas, when the resultant variable is a real and uninterrupted value, the problem is regression, here, a change in one variable is linked with a change in other variable (e.g., weight based on height). Some common examples of supervised ML in medicine is to perform pattern recognition over selected set of diagnosis by a cardiologist by interpretation of EKG and also from a chest X ray detection of lung module can be determined [5]. In medicine, for detection of risk in coronary heart disease, the best possible method adopted for analysis is Framingham Risk Score which is an example of supervised ML [6]. Risk models like above in medicine can guide in antithrombotic therapy in atrial fibrillation [7] and in hypertrophic cardiomyopathy for the implantation of electronic defibrillators [8].

Figure 1.3 The types of machine learning algorithm.

1.3.2 Unsupervised Learning

Such type of ML algorithm does not work upon labeled data and the machine learns from the dataset given and finds out the hidden pattern to make prediction about the output. It is further grouped into clustering and association; in clustering, the machine forms groups based on the behavior of the data, whereas association is a rule-based ML to discover relation between variables of large datasets. Precision medicine initiative is used to perform unsupervised learning problems in medicine [9]. How unsupervised learning can be applied in pathophysiologic mechanism to redefine the inherent heterogeneity in complex multi-factorial diseases, for instance, in cardiac disease like myocarditis. To apply the mechanism, inexplicable acute systolic heart failure is required and performed with myocardial biopsies to identify similar pattern between cellular compositions which will, in return, guide the therapist accordingly. Albiet the same technique to identify a subtype of asthma which responded to IL-13 [10, 11] is adopted.

1.3.3 Semi-Supervised Learning

It is a combination of supervised and unsupervised learning. It uses a combination of small portion of labeled data and massive collection of unlabeled to improve the prediction. The algorithm has ability to learn how to react on a particular situation based on the environment. The main aim of this method is to improve classification performance. This method is highly applicable in the healthcare sector when labeled data is not sufficiently available. It is applicable for classification of protein sequence typically due to the large size of DNA strands. Consistency enforcing strategy is mostly followed by this method [12]. It has been widely used for classification of medical images to reduce effort over labeling data [13–15]. Apart from this, for breast cancer analysis [16] and liver segmentation [17], a co-training mechanism has been applied.

1.3.4 Reinforcement Learning

This category of algorithm has no predefined data and the input depends upon the action taken by the agent and these actions are then recorded in the form of matrices so that it can serve as the memory to the agent. The agent explores the environment and collects data which is further used to get the output. In medicine, there are several instances of reinforcement learning (RL) application like for the development of therapy plan for lung cancer [18] and epilepsy [19]. Deep RL approach has been recently proposed for the therapy plan development on medical registry data [20] and also to learn treatment strategies for sepsis [21].

1.3.5 Deep Learning

Such algorithms has been widely used in the field of science for solving and analyzing problems related to healthcare by using different techniques for image analysis for obtaining information effectively. DL requires data to get information but, when combined with the medicinal data, makes the work complex for the researcher. Once the data is obtained, it can be applied accordingly in different field of medicine like prognosis, diagnosis, treatment, and clinical workflow. DL concept is used to build tool for skin cancer detection in dermatology [22]. Neural network training using DL method is applied for the computation of diabetic retinopathy severity by using the strength of pixels in fundus image [23].

1.4 Big Data in Healthcare

Sebastian Thrum, a computer scientist, once said, “Just as machines made human muscles a thousand times stronger, machines will make the human brain a thousand times more powerful.” This statement is the seed of what big data and ML is doing to healthcare today impacting human lives like never before opening new doors of possibilities and for much good.

In this digital era, massive amounts of data are being generated every moment, the digital universe which was about 130 exabytes (EB) in 2005 has expanded to about 40,000 EB in 2020 [24]. Such huge amount of data, known as big data, is a storehouse of critical information which can transform the way we provide healthcare services.

Figure 1.4 Sources of big data in healthcare.

Large amounts of data are generated in healthcare services, and these are from sources (Figure 1.4) as diverse as government agencies, patient portals, research studies, generic databases, electronic health records, public health records, wearable devices, smart phones, and many more. All these sources generate data in different formats which need to be not just merged but also made available instantly when needed; this is where big data and ML together empower healthcare services.

1.5 Application of Big Data in Healthcare

1.5.1 Electronic Health Records

This is one aspect of healthcare where the biggest challenge was assimilating data generated in different forms and different sources and to be able to replicate it instantly for analysis. This helps not just the healthcare provider by keeping record of medical history, tests, and allergies but also keeps the patient informed of any tests or appointments due.

1.5.2 Helping in Diagnostics

With medicine moving toward a more preventive and predictive science and also generalized toward a more individualized science, big data analytics has assumed a more integral role. Use of large volumes of data in fields like radiology, pathology, and oncology helps in arriving at an early diagnosis with computer-assisted devices increasing the accuracy of diagnosis and helping in early intervention thereby improving outcomes.

1.5.3 Preventive Medicine

Wearable devices are the new healthcare providers in today’s times and they have an increasingly important role to play in medicine in the times to come. These devices are the new physicians which continue to monitor an individual at all times having not just diagnostic but also predictive value. These devices can monitor many parameters and can connect with physicians even far away.

1.5.4 Precision Medicine

In robotic surgeries especially neurosurgeries and oncosurgeries where precision is of utmost importance, big data and ML are combined to deliver results that lead to better outcomes with lesser morbidity and mortality, which is a boon for all stakeholders.

1.5.5 Medical Research

Advancements in medical research have gathered pace with the availability of big data and ML. With all the data at hand, researchers are forever trying to find better cures. Varied data available for analysis helps in understanding why a particular treatment modality worked in a population group while it failed to bring the desired response in another, blood thinners working in one patient population and not working in another is an example. Researchers are using genetic information to personalise drug treatments [25].

Using algorithms, scientists have been able to identify molecules that activate a particular protein which is relevant to symptoms of Alzheimer’s disease and schizophrenia; this could lead to early development of drugs for the treatment of these as yet incurable diseases [26].

1.5.6 Cost Reduction

With the availability of medical data, the healthcare provider gets all the background information of his patient helping him to make decisions with less errors, resulting in lower costs for the patient and the healthcare system. Analysis of data from a particular patient population helps in deciding disease management strategies improving preventive care, thereby minimizing costs.

1.5.7 Population Health

Big data availability leads to optimal use of the available resources for the entire community. It provides insights as to which patient population is especially vulnerable to a particular illness, so that measures can be taken to lessen the impact of a disease whether communicable or non-communicable by scaling up the facilities needed for its management so that the impact of the disease can be contained.

1.5.8 Telemedicine

Doctors recommend telemedicine to patients for personalized treatment solutions to prevent readmissions, and data analytics can then be used to make assessments and predictions of the course and associated management adjustments.

1.5.9 Equipment Maintenance

With connectivity between healthcare infrastructures for seamless real-time operations, its maintenance to prevent breakdowns becomes the backbone of the entire system.

1.5.10 Improved Operational Efficiency

The big data helps us understand the admission, diagnosis, and records of utilization of resources, helping to understand the efficiency and productivity of the hospital facilities.

1.5.11 Outbreak Prediction

With the availability of data like temperature and rainfall, reported cases reasonable predictions can be made about the outbreak of vector borne diseases like malaria and encephalitis, saving lives.

1.6 Challenges for Big Data

The volumes of data generated from diverse sources need to be sorted into a cohesive format and then be constantly updated so that it can be shared between healthcare service providers addressing the relevant security concerns, which is the biggest challenge. Insights gained from big data help us in understanding the pooled data better, thereby helping in improving the outcomes and benefitting insurance providers by reducing fraud and false claims.

The healthcare industry needs skilled data analysts who can sift out relevant data, analyze it, and communicate it to the relevant decision makers.

1.7 Conclusion

This chapter provides an overview regarding better healthcare services with the help of ML and big data technology. It presents big data approaches to gather valuable medical records and further the application of ML algorithm. The implementation of ML tool in medicine shows more accurate result with less processing time. Big data will surely help in collecting and maintaining EHR (electronic health records) for better decision-making in future. This paper provides a systematic review to the researchers about better options in the field of healthcare using ML and big data (Figure 1.5). This paper has identified several application areas in healthcare services using ML and big data which can further improve the unresolved challenges. The traditional healthcare services will be greatly transformed by such technologies. ML will help in improving the relationship between the locals and service provider by providing better service in less time. It will help in keeping an eye on critical patients in real time and help them diagnose the disease and recommend further treatment.

Figure 1.5 Applications of big data in healthcare.

References

1. James, M., Michael, C., Brad, B., Jacques, B., Big Data: The Next Frontier for Innovation, Competition, and Productivity, McKinsey Global Institute, New York, NY, 2011.

2. Al-Jarrah, O.Y., Yoo, P.D., Muhaidat, S., Karagiannidis, G.K., Taha, K., Efficient machine learning for big data: A review. Big Data Res., 2, 3, 87–93, Sep. 2015.

3. Beyer, M.A. and Laney, D., The importance of ‘big data’: A definition, Gartner Research, Stamford, CT, USA, Tech. Rep. G00235055, 2012.

4. Xing, L., Krupinski, E.A., Cai, J., Artificial intelligence will soon change the landscape of medical physics research and practice. Med. Phys., 45, 5, 1791–1793, 2018.

5. Deo, R.C., Machine Learning in Medicine. Circulation, 132, 20, 1920–1930, 2015.

6. Kannel, W.B., Doyle, J.T., McNamara, P.M., Quickenton, P., Gordon, T., Precursors of sudden coronary death. Factors related to the incidence of sudden death. Circulation, 51, 606–613, 1975, [PubMed] [Google Scholar].

7. Lip, G.Y.H., Nieuwlaat, R., Pisters, R., Lane, D.A., Crijns, H.J.G.M., Refining clinical risk stratification for predicting stroke and thromboembolism in atrial fibrillation using a novel risk factor-based approach: the euro heart survey on atrial fibrillation. Chest, 137, 263–272, 2010, [PubMed] [Google Scholar].

8. O’Mahony, C., Jichi, F., Pavlou, M., Monserrat, L., Anastasakis, A., Rapezzi, C., Biagini, E., Gimeno, J.R., Limongelli, G., McKenna, W.J., Omar, R.Z., Elliott, P.M., A novel clinical risk prediction model for sudden cardiac death in hypertrophic cardiomyopathy (HCM risk-SCD). Eur. Heart J., 35, 2010–2020, 2014, [PubMed] [Google Scholar].

9. National Research Council (US), Committee on A Framework for Developing a New Taxonomy of Disease, in: Toward Precision Medicine: Building a Knowledge Network for Biomedical Research and a New Taxonomy of Disease, National Academies Press (US), Washington (DC), 2011, [PubMed] [Google Scholar].

10. Woodruff, P.G., Modrek, B., Choy, D.F., Jia, G., Abbas, A.R., Ellwanger, A., Koth, L.L., Arron, J.R., Fahy, J.V., T-helper type 2-driven inflammation defines major subphenotypes of asthma. Am. J. Respir. Crit. Care Med., 180, 388–395, 2009, [PMC free article] [PubMed] [Google Scholar].

11. Corren, J., Lemanske, R.F., Hanania, N.A., Korenblat, P.E., Parsey, M.V., Arron, J.R., Harris, J.M., Scheerens, H., Wu, L.C., Su, Z., Mosesova, S., Eisner, M.D., Bohen, S.P., Matthews, J.G., Lebrikizumab treatment in adults with asthma. N. Engl. J. Med., 365, 1088–1098, 2011, [PubMed] [Google Scholar].

12. Laine, S. and Aila, T., Temporal ensembling for semi-supervised learning, in: International Conference on Learning Representations, 2017.

13. Cheplygina, V., d. Bruijne, M., Pluim, J.P.W., Not-so-supervised: a survey of semi-supervised, multi-instance, and transfer learning in medical image analysis. Medical Image Analysis, 54, 280–296, 2019.

14. Bai, W., Oktay, O., Sinclair, M., Suzuki, H., Rajchl, M., Tarroni, G., Glocker, B., King, A., Matthews, P.M., Rueckert, D., Semisupervised learning for network-based cardiac mr image segmentation, in: