Advances in Computational Techniques for Biomedical Image Analysis: Methods and Applications

Advances in Computational Techniques for Biomedical Image Analysis
Copyright
Contents
List of contributors
1 Computational techniques in biomedical image analysis: overview
	1.1 Introduction
	1.2 Medical imaging modalities
		1.2.1 X-ray
		1.2.2 Computed tomography
		1.2.3 Magnetic resonance imaging
		1.2.4 Functional magnetic resonance imaging
		1.2.5 Magnetic resonance spectroscopy
		1.2.6 Ultrasound
		1.2.7 Elastography
		1.2.8 Nuclear medicine
			1.2.8.1 Scintigraphy
			1.2.8.2 Single photon emission computed tomography
			1.2.8.3 Positron emission tomography
			1.2.8.4 Nuclear magnetic resonance
		1.2.9 Optical imaging
		1.2.10 Fundus imaging
		1.2.11 Histopathological images
		1.2.12 Comparison and risks of medical imaging modalities
	1.3 Computational techniques in medical image analysis
		1.3.1 Image denoising
			1.3.1.1 Spatial domain technique
			1.3.1.2 Transform domain
		1.3.2 Image segmentation
		1.3.3 Image registration and fusion
		1.3.4 Medical image classification
		1.3.5 Medical image compression techniques for transmission
		1.3.6 Security in communication
	1.4 Discussion and conclusions
	References
	Further reading
2 Multimodal medical image fusion using deep learning
	2.1 Introduction
	2.2 Conventional multimodal medical image fusion system
		2.2.1 Imaging modalities
		2.2.2 Components of medical image fusion system
			2.2.2.1 Image decomposition and reconstruction methods
			2.2.2.2 Image fusion rules
			2.2.2.3 Image quality assessment
		2.2.3 Need of multimodal image fusion in medical sector
			2.2.3.1 Objectives of medical image fusion using multiple modalities
	2.3 Inspiration to use deep learning for image fusion
		2.3.1 Complexities associated with conventional approaches of image fusion
		2.3.2 Advantages of deep learning in image fusion
	2.4 Frequently employed deep learning models in the field of medical image fusion
		2.4.1 Convolutional neural networks
		2.4.2 Convolutional sparse representation
		2.4.3 Stacked auto encoders
	2.5 Review of deep learning-based image fusion techniques
	2.6 Conclusion
	References
3 Medical image fusion framework for neuro brain analysis
	3.1 Introduction
	3.2 Fractional Fourier transform
	3.3 Material and method
		3.3.1 Gray and color neuro images
		3.3.2 Neurological data
		3.3.3 Fractional process domain
	3.4 Result and discussion
	3.5 Conclusion
	References
4 Automated detection of intracranial hemorrhage in noncontrast head computed tomography
	4.1 Introduction
	4.2 Intracranial hemorrhage
	4.3 Neuroimaging techniques for intracranial hemorrhage
		4.3.1 Noncontrast computed tomography
		4.3.2 Magnetic resonance imaging
		4.3.3 Computed tomography angiography
		4.3.4 Dual energy computed tomography
	4.4 Presentation of intracranial hemorrhage on noncontrast head computed tomography and need for automation
	4.5 Automation techniques in medical imaging
	4.6 Traditional machine learning
		4.6.1 Preprocessing
		4.6.2 Feature extraction and Selection
		4.6.3 Classification
	4.7 Challenges in using traditional methods
	4.8 Deep learning
		4.8.1 Visualization for deep learning
	4.9 Automated detection techniques in intracranial hemorrhage
		4.9.1 Preprocessing
			4.9.1.1 Resampling
			4.9.1.2 Contrast enhancement and windowing
			4.9.1.3 Skull and background removal
			4.9.1.4 Noise removal
			4.9.1.5 Composite image creation
			4.9.1.6 Identification of hemorrhage region of interest
		4.9.2 Traditional machine learning
		4.9.3 Deep learning
			4.9.3.1 Two-dimensional deep convolutional neural networks
			4.9.3.2 Three-dimensional deep convolutional neural networks
			4.9.3.3 Hybrid deep convolutional neural networks
	4.10 Clinical applications
	4.11 Discussion and conclusion
	References
5 Segmentation techniques for the diagnosis of intervertebral disc diseases
	5.1 Introduction
	5.2 Intervertebral disc segmentation techniques
	5.3 Herniated intervertebral disc segmentation techniques
	5.4 Challenges in the segmentation of the vertebra and intervertebral discs
		5.4.1 Challenges of spinal magnetic resonance imaging
		5.4.2 Challenges in the segmentation of the vertebra and intervertebral discs
	5.5 Conclusion and future work
	References
6 Heartbeat sound classification using Mel-frequency cepstral coefficients and deep convolutional neural network
	6.1 Introduction
	6.2 Literature review
	6.3 Mel-frequency cepstral coefficients
	6.4 Convolution neural network
	6.5 Heartbeat sound database
	6.6 Experiments
		6.6.1 Experiment 1: Heartbeat sound classification with discrete cosine transform basis type-1
		6.6.2 Experiment 2: Heartbeat sound classification with discrete cosine transform basis type-2
		6.6.3 Experiment 3: Heartbeat sound classification with discrete cosine transform basis type-3
	6.7 Conclusion
	References
7 Comparative analysis of classification techniques for brain magnetic resonance imaging images
	7.1 Introduction
	7.2 Literature review
	7.3 Methodology
		7.3.1 Gray level co-occurrence matrix
		7.3.2 Support vector machine
		7.3.3 Self-organizing maps
		7.3.4 Fuzzy c-means clustering
		7.3.5 Probabilistic neural network
		7.3.6 Convolution neural network
	7.4 Comparative analysis of various approaches
	7.5 Conclusion
	References
8 Hybrid feature selection-based feature fusion for liver disease classification on ultrasound images
	8.1 Introduction
	8.2 Method
		8.2.1 Feature extraction
			8.2.1.1 Feature normalization
		8.2.2 Feature selection
			8.2.2.1 ReliefF
			8.2.2.2 Sequential forward selection
			8.2.2.3 Hybrid feature selection
			8.2.2.4 Feature fusion
		8.2.3 Classification methods
	8.3 Experiments and results
		8.3.1 Experiment 1: Effectiveness of texture features and feature dimensionality reduction
		8.3.2 Experiment 2: Effectiveness of feature fusion
		8.3.3 Experiment 3: Effectiveness of proposed feature selection strategy
	8.4 Discussions
	8.5 Conclusion
	References
9 Discrete cosine transform–based compressive sensing recovery strategies in medical imaging
	9.1 Introduction
	9.2 Literature review
	9.3 Methodology
		9.3.1 Compressive sensing
		9.3.2 CS recovery algorithms
			9.3.2.1 L1-magic
			9.3.2.2 Orthogonal matching pursuit
			9.3.2.3 Compressive sampling matching pursuit
			9.3.2.4 CVX
		9.3.3 Proposed weighted compressive sensing
		9.3.4 Performance metrics
	9.4 Results and discussion
		9.4.1 Comparative study using traditional recovery schemes
		9.4.2 Comparative study using proposed weighting-based recovery methods
	9.5 Conclusions
	References
10 Segmentation-based compression techniques for medical images
	10.1 Introduction
	10.2 Research and developments in region of interest coding
	10.3 Classification of segmentation-based coding techniques
		10.3.1 Unsupervised/region-based image compression
		10.3.2 Supervised/content-based image compression
	10.4 Comparative analysis of segmentation techniques
	10.5 New trends
		10.5.1 Deep learning
		10.5.2 Visual saliency Mmodels
	10.6 Challenges and future scope
	References
11 Systematic survey of compression algorithms in medical imaging
	11.1 Introduction
	11.2 Modalities of medical imaging
	11.3 File formats in medical imaging
		11.3.1 Parameters
		11.3.2 Formats
	11.4 Different compression techniques
		11.4.1 Lossless compression methods
			11.4.1.1 Joint photographic experts group lossless
			11.4.1.2 Stationary wavelet transform-based lossless compression
		11.4.2 Lossy compression methods
			11.4.2.1 Wavelet transform-based lossy compression
			11.4.2.2 Joint photographic experts group lossy compression
		11.4.3 Advanced compression method
			11.4.3.1 Region of interest-based compression algorithm
			11.4.3.2 Joint photographic experts group 2000
			11.4.3.3 Joint photographic experts group extended range
			11.4.3.4 Ripplet transform-based compression
	11.5 Summary and discussion
	11.6 Conclusion and future directions
	Disclosure statement
	References
12 Multilevel medical image encryption for secure communication
	12.1 Introduction
	12.2 Related work
	12.3 Materials and methods
		12.3.1 Permutation via Arnold cat map
		12.3.2 Proposed multiple chaotic map transformation
	12.4 Results and analysis
		12.4.1 Histogram analysis
		12.4.2 Image entropy
		12.4.3 Line profile
		12.4.4 Correlation coefficient
		12.4.5 Key space of the proposed cryptosystem
	12.5 Conclusion
	References
13 A modified digital signature algorithm to improve the biomedical image integrity in cloud environment
	13.1 Introduction
		13.1.1 Challenges
	13.2 Literature survey
	13.3 Proposed work
		13.3.1 Base formula
		13.3.2 Adler32
		13.3.3 Modified digital signature algorithm
		13.3.4 Mathematical proof
			13.3.4.1 Key generation part
			13.3.4.2 Adler32 hash function
			13.3.4.3 Caesar cipher table
			13.3.4.4 Signing formula
			13.3.4.5 Verification formula
			13.3.4.6 Comparison
			13.3.4.7 Values to signature table
	13.4 Experimental results
	13.5 Conclusion and future works
	References
14 Medical imaging security and forensics: a systematic literature review
	14.1 Introduction
	14.2 Related work
	14.3 Feasible study to find essential factors
		14.3.1 Handling of watermark
		14.3.2 Model quality of imaging
			14.3.2.1 Intelligent
			14.3.2.2 Transmissive
			14.3.2.3 Multimodal imaging
		14.3.3 Tools for proving image authenticity
		14.3.4 Digital imaging techniques to different forensic applications
		14.3.5 Classification of digital forensics
		14.3.6 Image forgery detection
	14.4 Comparative analysis and discussions
	14.5 Future research challenges
	14.6 Conclusions
	References
Index