Deep learning (DL) is a method of machine learning, running over artificial neural networks, that uses multiple layers to extract high-level features from large amounts of raw data. DL methods apply levels of learning to transform input data into more abstract and composite information. Handbook for Deep Learning in Biomedical Engineering: Techniques and Applications gives readers a complete overview of the essential concepts of DL and its applications in the field of biomedical engineering. DL has been rapidly developed in recent years, in terms of both methodological constructs and practical applications. DL provides computational models of multiple processing layers to learn and represent data with higher levels of abstraction. It is able to implicitly capture intricate structures of large-scale data and is ideally suited to many of the hardware architectures that are currently available. The ever-expanding amount of data that can be gathered through biomedical and clinical information sensing devices necessitates the development of machine learning and artificial intelligence techniques such as DL and convolutional neural networks to process and evaluate the data. Some examples of biomedical and clinical sensing devices that use DL include computed tomography (CT), magnetic resonance imaging (MRI), ultrasound, single photon emission computed tomography (SPECT), positron emission tomography (PET), magnetic particle imaging, electroencephalography/magnetoencephalography (EE/MEG), optical microscopy and tomography, photoacoustic tomography, electron tomography, and atomic force microscopy. Handbook for Deep Learning in Biomedical Engineering: Techniques and Applications provides the most complete coverage of DL applications in biomedical engineering available, including detailed real-world applications in areas such as computational neuroscience, neuroimaging, data fusion, medical image processing, neurological disorder diagnosis for diseases such as Alzheimer’s, attention deficit hyperactivity disorder (ADHD), and autism spectrum disorder (ASD), tumor prediction, and translational multimodal imaging analysis.
Author(s): Valentina Emilia Balas, Brojo Kishore Mishra, Raghvendra Kumar
Publisher: Academic Press
Year: 2020
Language: English
Pages: 320
City: New York
HANDBOOK OF DEEPLEARNING IN BIOMEDICAL ENGINEERING
Copyright
Contributors
About the editors
Preface
Key features
About the book
1 . Congruence of deep learning in biomedical engineering: future prospects and challenges
1. Introduction
1.1 SqueezeNet (image classification)
1.1.1 Strategies of architectural design
2. Fire module
3. Background study
3.1 Need of security
3.1.1 Types of security methods
3.1.1.1 Steganography
3.1.1.2 Watermarking
3.1.1.3 Cryptography
3.2 Advantages of steganography over cryptography
3.2.1 Resolution of steganography
3.2.2 Types of steganography
3.2.3 Image steganography
3.2.4 Image steganography method
3.3 Steganography techniques
3.3.1 Spatial domain technique
3.3.1.1 Least significant bit technique
3.3.2 Transform domain technique
3.4 Advantages of transform domain over spatial domain
3.5 Related study
3.5.1 DWT based
3.5.2 IWT based
3.6 Advantages of IWT over DWT
4. Study of various types of model
5. Proposed method by the authors
5.1 2D Haar wavelet transform
5.2 Huffman encoding technique
5.3 Embedding algorithm
6. Conclusion and future work
References
2 . Deep convolutional neural network in medical image processing
1. Introduction
2. Medical image analysis
2.1 Segmentation
2.2 Detection or diagnosis by computer-aided system
2.3 Detection and classification of abnormality
2.4 Registration
3. Convolutional neural network and its architectures
3.1 Architectures of deep convolutional neural network
3.1.1 General classification architectures
3.1.2 Multistream architectures
3.1.3 Segmentation architectures
4. Application of deep convolutional neural network in medical image analysis
4.1 Brain
4.2 Eye
4.3 Breast
4.4 Chest
4.5 Cardiac
4.6 Abdomen
5. Critical discussion: inferences for future work and limitations
6. Conclusion
References
3 . Application, algorithm, tools directly related to deep learning
1. Introduction
2. Tools used in deep learning
2.1 TensorFlow
2.1.1 Tensor data structure
2.1.2 Rank
2.1.3 Shape
2.1.4 Type
2.1.5 One-dimensional Tensor
2.1.6 Two-dimensional Tensor
2.2 Keras
2.2.1 Backend in Keras
2.2.2 Installing keras: Amazon Web Service
2.3 CAFFE
2.3.1 The main features of CAFFE
2.4 Torch tool
2.5 Theano
3. Algorithms
3.1 Deep belief networks
3.1.1 Architecture of Deep belief network
3.1.2 Working of deep belief network
3.2 Convolutional neural network
3.2.1 Input image
3.2.2 Convolution layer—the kernel
3.3 Recurrent neural network
3.3.1 How recurrent neural network works
3.4 Long short-term memory networks
3.4.1 Structure of long short-term memory
3.5 Stacked autoencoders
3.6 Deep Boltzmann Machine
4. Applications of deep learning
5. Conclusion
References
4 . A critical review on using blockchain technology in education domain
1. Introduction
2. Consortium blockchain and its suitability for e-governance
3. Consensus
3.1 Proof approaches
3.2 Vote-based approaches
3.3 Directed acyclic graph approaches
4. Attacks on blockchain
5. Blockchain in education domain
6. Scalability challenges
7. Security challenges
8. Conclusion
References
Further reading
5 . Depression discovery in cancer communities using deep learning
1. Introduction
2. Related work
2.1 Lexicon-based approaches
2.2 Machine learning–based approaches
2.2.1 Supervised machine learning
2.2.2 Sentiment analysis using supervised machine learning
2.2.3 Metaclassifiers
2.3 Hybrid approaches
2.4 Other techniques
2.5 Sentiment analysis for online depression detection
3. Proposed system architecture
3.1 Continuous bag of words
3.1.1 Skip-gram model
3.1.2 Word embedding optimization
4. Models
4.1 Convolutional neural network
4.1.1 Variants of convolutional neural network
4.2 Recurrent neural network
4.3 Long short-term memory
4.3.1 Bidirectional long short-term memory
5. Conclusion
References
Further reading
6 . Plant leaf disease classification based on feature selection and deep neural network
1. Introduction
2. Literature review
2.1 Plant diseases recognition using convolutional neural networks
2.2 Plant diseases recognition with artificial neural network
2.3 Feature selection
3. Our proposed framework
3.1 Data set
3.2 Image preprocessing
3.3 Convolutional neural network
3.3.1 AlexNet (2012)
3.3.2 VGG16 (2014)
3.3.3 ResNet (2015)
4. Results
4.1 Conventional models
4.2 Models with transfer learning
4.3 Multilayer perceptron approach
4.3.1 Feature extraction
4.3.2 Feature selection
4.3.2.1 Particle swarm optimization
4.3.2.2 Gray wolf optimization
4.3.2.3 Proposed adaptive particle–gray wolf optimization heuristic
4.3.2.4 Wrapper-based adaptive particle–gray wolf optimization
5. Conclusion
References
Further reading
7 . Early detection and diagnosis using deep learning
1. Introduction
1.1 Introduction to deep learning
1.1.1 Applications
1.1.2 Challenges faced by deep learning
1.2 Introduction to biomedical engineering
1.2.1 Branches of biomedical engineering
1.2.2 Challenges faced by biomedical engineering
2. Diagnostics using deep learning
2.1 Motivation for use of deep learning in diagnostics
2.2 Challenges and solutions
2.2.1 Retroactive versus forthcoming trainings
2.2.2 Metric cannot be used for medical applicability
2.2.3 Trouble associating dissimilar algorithms
2.2.4 Hominoid barricades to artificial intelligence acceptance in medical sector
2.2.5 Vulnerability to confrontational occurrence or management
2.3 Future of diagnostics using deep learning
3. Early detection of diseases using deep learning
3.1 Rheumatic diseases
3.2 Alzheimer's disease
3.3 Autism spectrum disorder
3.4 Attention deficit hyperactivity disorder
4. Conclusion and further advancements
References
8 . A review on plant diseases recognition through deep learning
1. Introduction
2. Plant diseases
3. Traditional methods to treat plant diseases
3.1 Serological assays
3.1.1 Modern serological methods
3.1.1.1 Enzyme-linked immunosorbent assay
3.1.1.2 Dot blot immunobinding assay
3.1.1.3 Tissue blotting immunoassay
3.2 Nuclei acid–based methods
3.2.1 Polymerase chain reaction
3.2.2 Restriction fragment length polymorphisms
3.2.3 Amplified fragment length polymorphism
4. Innovative detection method
4.1 Lateral flow microarrays
4.2 Methods based on the analysis of volatile compounds as biomarkers
5. Remote sensing of plant diseases
5.1 Detection of plant impairment using remote sensing systems
5.2 Remote sensing systems for monitoring pests and diseases
5.3 Visible and short-wave infrared monitoring systems
5.4 Fluorescence and thermal sensors
6. Plant disease detection by well-known deep learning architectures
6.1 Evolution of Deep learning
6.2 Without visualization technique
6.3 Visualization techniques
6.4 Hyperspectral imaging with deep learning models
7. Conclusions
References
9 . Applications of deep learning in biomedical engineering
1. Introduction
2. Biomedical engineering
3. Deep learning
4. Most popular deep neural networks architectures used in biomedical applications
5. Convolutional neural network
6. Convolution layer
7. Pooling layer
8. Fully convolutional layer
9. Applications of convolutional neural network in biomedicine
10. Recurrent neural network
11. Applications of recurrent neural network in biomedicine
12. Generative adversarial networks
12.1 Generator network
12.2 Discriminator network
13. Applications of generative adversarial network in biomedicine
14. Deep belief network
15. Pretraining stage
16. Fine-tuning stage
17. Applications of deep learning in biomedicine
18. Biomedical image analysis
19. Image detection and recognition
20. Image acquisition and image interpretation
21. Image segmentation
22. Cytopathology and histopathology
23. Brain, body, and machine interface
23.1 Brain–machine interface
24. Classification of the brain–machine interfaces
25. Invasive techniques
26. Noninvasive techniques
27. Body–machine interface
28. Drug infusion system
29. Rehabilitation system
30. Diseases diagnosis
31. Omics
32. Around the genome
33. Protein-binding prediction
34. DNA–RNA-binding proteins
35. Gene expression
36. Alternative splicing
37. Gene expression prediction
38. Genomic sequencing
39. Around the protein
40. Protein Structure Prediction
41. Protein secondary structure prediction
41.1 Protein tertiary structure prediction
41.2 Protein quality assessment
41.3 Protein loop modeling and disorder prediction
42. Protein Interaction Prediction
42.1 Protein–protein interactions
42.2 Drug–target interactions
43. Public and medical health management
44. Conclusion
References
10 . Deep neural network in medical image processing
1. Literature review
2. Digital image and computer vision
2.1 Introduction
2.2 What is an image?
2.3 Digital representation of an image
2.4 Medical image formats
2.5 Steps in image processing
2.6 Machine learning and its types
2.7 Unlabeled data set
2.8 Labeled data set
2.9 Supervised learning
2.10 Unsupervised learning
2.11 Reinforcement learning
2.12 Artificial neural network
3. Deep learning
3.1 Deep learning architectures
4. Segmentation techniques in image processing
4.1 Different approaches for segmentation
4.2 Edge-based segmentation methods
4.3 Threshold segmentation
5. Conclusion
References
Index
