Medical Image Analysis

ATAM P. DHAWAN, PHD, is Distinguished Professor in the Electrical and Computer Engineering Department at New Jersey Institute of Technology. He teaches courses in biomedical engineering and has supervised approximately fifty graduate students, including twenty-one PhD students. Dr. Dhawan is a Fellow of the IEEE and the recipient of numerous national and international awards. He has published more than 200 research articles in refereed journals, conference proceedings, and edited books. Dr. Dhawan has chaired numerous study sections and review panels for the National Institutes of Health in biomedical computing and medical imaging and health informatics. His current research interests are medical imaging, multi-modality medical image analysis, multi-grid image reconstruction, wavelets, genetic algorithms, neural networks, adaptive learning, and pattern recognition.

• Preface to the Second Edition xiiiChapter 1 Introduction 11.1. Medical Imaging: A Collaborative Paradigm 21.2. Medical Imaging Modalities 31.3. Medical Imaging: from Physiology to Information Processing 61.3.1 Understanding Physiology and Imaging Medium 61.3.2 Physics of Imaging 71.3.3 Imaging Instrumentation 71.3.4 Data Acquisition and Image Reconstruction 71.3.5 Image Analysis and Applications 81.4. General Performance Measures 81.4.1 An Example of Performance Measure 101.5. Biomedical Image Processing and Analysis 111.6. Matlab Image Processing Toolbox 141.6.1 Digital Image Representation 141.6.2 Basic MATLAB Image Toolbox Commands 161.7. Imagepro Interface in Matlab Environment and Image Databases 191.7.1 Imagepro Image Processing Interface 191.7.2 Installation Instructions 201.8. Imagej and Other Image Processing Software Packages 201.9. Exercises 211.10. References 221.11. Definitions 22Chapter 2 Image Formation232.1. Image Coordinate System 242.1.1 2-D Image Rotation 252.1.2 3-D Image Rotation and Translation Transformation 262.2. Linear Systems 272.3. Point Source and Impulse Functions 272.4. Probability and Random Variable Functions 292.4.1 Conditional and Joint Probability Density Functions 302.4.2 Independent and Orthogonal Random Variables 312.5. Image Formation 322.5.1 PSF and Spatial Resolution 352.5.2 Signal-to-Noise Ratio 372.5.3 Contrast-to-Noise Ratio 392.6. Pin-hole Imaging 392.7. Fourier Transform 402.7.1 Sinc Function 432.8. Radon Transform 442.9. Sampling 462.10. Discrete Fourier Transform 502.11. Wavelet Transform 522.12. Exercises 602.13. References 62Chapter 3 Interaction of Electromagnetic Radiation with Matter in Medical Imaging 653.1. Electromagnetic Radiation 653.2. Electromagnetic Radiation for Image Formation 663.3. Radiation Interaction with Matter 673.3.1 Coherent or Rayleigh Scattering 673.3.2 Photoelectric Absorption 683.3.3 Compton Scattering 693.3.4 Pair Production 693.4. Linear Attenuation Coefficient 703.5. Radiation Detection 703.5.1 Ionized Chambers and Proportional Counters 703.5.2 Semiconductor Detectors 723.5.3 Advantages of Semiconductor Detectors 733.5.4 Scintillation Detectors 733.6. Detector Subsystem Output Voltage Pulse 763.7. Exercises 783.8. References 78Chapter 4 Medical Imaging Modalities: X-Ray Imaging 794.1. X-Ray Imaging 804.2. X-Ray Generation 814.3. X-Ray 2-D Projection Imaging 844.4. X-Ray Mammography 864.5. X-Ray CT 884.6. Spiral X-Ray CT 924.7. Contrast Agent, Spatial Resolution, and SNR 954.8. Exercises 964.9. References 97Chapter 5 Medical Imaging Modalities: Magnetic Resonance Imaging 995.1. MRI Principles 1005.2. MR Instrumentation 1105.3. MRI Pulse Sequences 1125.3.1 Spin-Echo Imaging 1145.3.2 Inversion Recovery Imaging 1185.3.3 Echo Planar Imaging 1195.3.4 Gradient Echo Imaging 1235.4. Flow Imaging 1255.5. fMRI 1295.6. Diffusion Imaging 1305.7. Contrast, Spatial Resolution, and SNR 1355.8. Exercises 1375.9. References 138Chapter 6 Nuclear Medicine Imaging Modalities 1396.1. Radioactivity 1396.2. SPECT 1406.2.1 Detectors and Data Acquisition System 1426.2.2 Contrast, Spatial Resolution, and Signal-to-Noise Ratio in SPECT Imaging 1456.3. PET 1486.3.1 Detectors and Data Acquisition Systems 1506.3.2 Contrast, Spatial Resolution, and SNR in PET Imaging 1506.4. Dual-Modality Spect–CT and PET–CT Scanners 1516.5. Exercises 1546.6. References 155Chapter 7 Medical Imaging Modalities: Ultrasound Imaging 1577.1. Propagation of Sound in a Medium 1577.2. Reflection and Refraction 1597.3. Transmission of Ultrasound Waves in a Multilayered Medium 1607.4. Attenuation 1627.5. Ultrasound Reflection Imaging 1637.6. Ultrasound Imaging Instrumentation 1647.7. Imaging with Ultrasound: A-Mode 1667.8. Imaging with Ultrasound: M-Mode 1677.9. Imaging with Ultrasound: B-Mode 1687.10. Doppler Ultrasound Imaging 1697.11. Contrast, Spatial Resolution, and SNR 1707.12. Exercises 1717.13. References 172Chapter 8 Image Reconstruction 1738.1. Radon Transform and Image Reconstruction 1748.1.1 The Central Slice Theorem 1748.1.2 Inverse Radon Transform 1768.1.3 Backprojection Method 1768.2. Iterative Algebraic Reconstruction Methods 1808.3. Estimation Methods 1828.4. Fourier Reconstruction Methods 1858.5. Image Reconstruction in Medical Imaging Modalities 1868.5.1 Image Reconstruction in X-Ray CT 1868.5.2 Image Reconstruction in Nuclear Emission Computed Tomography: SPECT and PET 1888.5.2.1 A General Approach to ML–EM Algorithms 1898.5.2.2 A Multigrid EM Algorithm 1908.5.3 Image Reconstruction in Magnetic Resonance Imaging 1928.5.4 Image Reconstruction in Ultrasound Imaging 1938.6. Exercises 1948.7. References 195Chapter 9 Image Processing and Enhancement 1999.1. Spatial Domain Methods 2009.1.1 Histogram Transformation and Equalization 2019.1.2 Histogram Modification 2039.1.3 Image Averaging 2049.1.4 Image Subtraction 2049.1.5 Neighborhood Operations 2059.1.5.1 Median Filter 2079.1.5.2 Adaptive Arithmetic Mean Filter 2079.1.5.3 Image Sharpening and Edge Enhancement 2089.1.5.4 Feature Enhancement Using Adaptive Neighborhood Processing 2099.2. Frequency Domain Filtering 2129.2.1 Wiener Filtering 2139.2.2 Constrained Least Square Filtering 2149.2.3 Low-Pass Filtering 2159.2.4 High-Pass Filtering 2179.2.5 Homomorphic Filtering 2179.3. Wavelet Transform for Image Processing 2209.3.1 Image Smoothing and Enhancement Using Wavelet Transform 2239.4. Exercises 2269.5. References 228Chapter 10 Image Segmentation 22910.1. Edge-Based Image Segmentation 22910.1.1 Edge Detection Operations 23010.1.2 Boundary Tracking 23110.1.3 Hough Transform 23310.2. Pixel-Based Direct Classification Methods 23510.2.1 Optimal Global Thresholding 23710.2.2 Pixel Classification Through Clustering 23910.2.2.1 Data Clustering 23910.2.2.2 k-Means Clustering 24110.2.2.3 Fuzzy c-Means Clustering 24210.2.2.4 An Adaptive FCM Algorithm 24410.3. Region-Based Segmentation 24510.3.1 Region-Growing 24510.3.2 Region-Splitting 24710.4. Advanced Segmentation Methods 24810.4.1 Estimation-Model Based Adaptive Segmentation 24910.4.2 Image Segmentation Using Neural Networks 25410.4.2.1 Backpropagation Neural Network for Classification 25510.4.2.2 The RBF Network 25810.4.2.3 Segmentation of Arterial Structure in Digital Subtraction Angiograms 25910.5. Exercises 26110.6. References 262Chapter 11 Image Representation, Analysis, and Classification 26511.1. Feature Extraction and Representation 26811.1.1 Statistical Pixel-Level Features 26811.1.2 Shape Features 27011.1.2.1 Boundary Encoding: Chain Code 27111.1.2.2 Boundary Encoding: Fourier Descriptor 27311.1.2.3 Moments for Shape Description 27311.1.2.4 Morphological Processing for Shape Description 27411.1.3 Texture Features 28011.1.4 Relational Features 28211.2. Feature Selection for Classification 28311.2.1 Linear Discriminant Analysis 28511.2.2 PCA 28811.2.3 GA-Based Optimization 28911.3. Feature and Image Classification 29211.3.1 Statistical Classification Methods 29211.3.1.1 Nearest Neighbor Classifier 29311.3.1.2 Bayesian Classifier 29311.3.2 Rule-Based Systems 29411.3.3 Neural Network Classifiers 29611.3.3.1 Neuro-Fuzzy Pattern Classification 29611.3.4 Support Vector Machine for Classification 30211.4. Image Analysis and Classification Example: “Difficult-To-Diagnose” Mammographic Microcalcifications 30311.5. Exercises 30611.6. References 307Chapter 12 Image Registration 31112.1. Rigid-Body Transformation 31412.1.1 Affine Transformation 31612.2. Principal Axes Registration 31612.3. Iterative Principal Axes Registration 31912.4. Image Landmarks and Features-Based Registration 32312.4.1 Similarity Transformation for Point-Based Registration 32312.4.2 Weighted Features-Based Registration 32412.5. Elastic Deformation-Based Registration 32512.6. Exercises 33012.7. References 331Chapter 13 Image Visualization 33513.1. Feature-Enhanced 2-D Image Display Methods 33613.2. Stereo Vision and Semi-3-D Display Methods 33613.3. Surface- and Volume-Based 3-D Display Methods 33813.3.1 Surface Visualization 33913.3.2 Volume Visualization 34413.4. VR-Based Interactive Visualization 34713.4.1 Virtual Endoscopy 34913.5. Exercises 34913.6. References 350Chapter 14 Current and Future Trends in Medical Imaging and Image Analysis 35314.1. Multiparameter Medical Imaging and Analysis 35314.2. Targeted Imaging 35714.3. Optical Imaging and Other Emerging Modalities 35714.3.1 Optical Microscopy 35814.3.2 Optical Endoscopy 36014.3.3 Optical Coherence Tomography 36014.3.4 Diffuse Reflectance and Transillumination Imaging 36214.3.5 Photoacoustic Imaging: An Emerging Technology 36314.4. Model-Based and Multiscale Analysis 36414.5. References 366Index 503