Infrared thermography is a fast and non-invasive technology that provides a map of the temperature distribution on the body's surface. This book provides a description of designing and developing a computer-assisted diagnosis (CAD) system based on thermography for diagnosing such common ailments as rheumatoid arthritis (RA), diabetes complications, and fever. It also introduces applications of machine-learning and deep-learning methods in the development of CAD systems.

Key Features:

• Covers applications of various image processing techniques in thermal imaging applications for the diagnosis of different medical conditions
• Describes the development of a computer diagnostics system (CAD) based on thermographic data
• Discusses deep-learning models for accurate diagnosis of various diseases
• Includes new aspects in rheumatoid arthritis and diabetes research using advanced analytical tools
• Reviews application of feature fusion algorithms and feature reduction algorithms for accurate classification of images

This book is aimed at researchers and graduate students in biomedical engineering, medicine, image processing, and CAD.

Dr. U. Snekhalatha is currently working as an Associate professor in the Department of Biomedical Engineering, College of Engineering and Technology, SRM Institute of Science and Technology, Kattankulathur. She pursued her Doctorate in Biomedical Engineering at SRMIST (2015). Her area of interest includes biomedical signal processing, medical image processing, biomedical instrumentation, machine learning and deep learning techniques. She published 21 research articles in a reputed peer reviewed international journals with good impact factor indexed in science citation index, medline and Pubmed etc. She has published 22 Scopus indexed international journals and 25 technical papers published in various national and IEEE International conferences indexed in IEEE Xplore digital library and other international conferences. She has filed 4 Indian patents in which all are in the published stage. She is a life member of various professional societies such as Biomedical society of India, IEI, IEANG, IRED, ISTE and ISCA. She is currently serving as a reviewer in various reputed peer reviewed international journals. Dr. Palani Thanaraj Krishnan is currently working as Assistant Professor in the Department of Electronics & Instrumentation Engineering, St. Joseph's College of Engineering, Chennai. He has completed his PhD in 2018 in the Faculty of Information and Communication Engineering from Anna University, Chennai. His research areas include Image Processing, Advanced signal processing, Image segmentation, Machine learning, and Deep learning. He has developed deep learning algorithms for performing image classification of medical images for disease diagnosis. He has published his works in many reputed and refereed journals indexed in Web of Science and Scopus. Prof Dr. med Kurt Ammer was certified as a general medical practitioner in 1978, a consultant for physical medicine and rehabilitation in 1989, and a consultant for physical medicine and rehabilitation (rheumatology) in 1994. He was senior researcher at the Ludwig Boltzmann Research Unit for Physical Diagnostics, Austria, between 1988 and 2004. From 1985 till his retirement in early 2013, he was vice director of the Institute of Physical Medicine and Rehabilitation at the Hanusch hospital in Vienna. He got involved in medical thermography in 1988, and was appointed as secretary and treasurer of the European Association of Thermology in 1990, and currently serves as the EAT treasurer. Since 2002, he has been appointed as external professor at the Medical Imaging Research Unit, University of South Wales, Pontypridd, UK. His research interests focus on rehabilitation medicine and the application and standardization of thermal imaging in medicine.

1. Fundamentals of Infrared Thermal Imaging 1.1 Basics of Thermometry 1.2 Thermal Properties of Biological Tissues 1.3 Thermal Signals Associated with Physiological Functions 1.3.1 Temperature signal of respiratio 1.3.2 Thermal heart rate signal 1.3.3 Identification of sweat 1.3.4 Facial muscle activation signals 1.4 Examples of Recent Applications of Infrared Thermal Imaging in Medicine 1.4.1 Rheumatoid Arthritis 1.4.2 Diabetic Mellitus 1.4.3 Dermatology 1.4.4 Dental Disorders 1.4.5 Complex regional pain syndrome 1.4.6 Fever screening 1.5 Summary 2. Protocol for Standardized data collection in humans 2.1 Back ground and International Guidelines 2.2 Thermal Camera Performance Check 2.3 Ambient Temperature control 2.3.1 Choice of ambient temperature 2.3.2 Structure of the examination space 2.4 Subject Selection 2.4.1 Procedures performed during the examination 2.4.2 Behavior prior to the thermal imaging 2.4.3 Data checking at the beginning of the imaging session 2.5 Patient Position and Image acquisition 2.5.1 Static or dynamic thermal imaging 2.6 Thermal image analysis 3. Basic Approaches of Artificial Intelligence and Machine learning in Thermal Image Processing 3.1 Image source 3.2 Data Transfer from camera to the analyzing software 3.3 Preprocessing 3.3.1 Thermal Image Color Modes 3.3.2 Image Enhancement 3.3.2.1 Image Smoothening and Sharpening 3.3.2.2 Histogram Equalization 3.3.3 Multispectral Images 3.3.4 Image Registration 3.3.5 Image Fusion 3.4 Segmentation Techniques 3.4.1 Clustering based Segmentation Algorithms 3.4.1.1 K-Means Clustering 3.4.1.2 Hierarchical Clustering 3.4.1.3 Divisive Clustering 3.4.1.4 Density-Based Clustering 3.4.1.5 Fuzzy C-Means Clustering 3.4.1.6 Neutrosophic C-Means Clustering 3.4.1.7 Mean Shift Clustering 3.4.2 Threshold-based Segmentation 3.4.3 Region-based Segmentation Algorithms 3.4.3.1 Region Growing 3.4.3.2 Region Merging 3.4.4 Active Contour 3.4.5 Edge Detection for Image Segmentation 3.4.5.1 Roberts Edge Detection 3.4.5.2 Sobel Edge Detection 3.4.5.3 Prewitt Edge Detection 3.4.5.4 Marr-Hildreth Edge Detection 3.4.5.5 Canny Edge Detection 3.4.6 Watershed Segmentation 3.4.7 Deep Learning Models For Image Segmentation 3.5 Artifact Suppression 3.6 Feature Extraction and Classification 3.6.1 Gray Level Co-Occurrence Matrix 3.6.2 Speeded Up Robust Features 3.6.3 Gray Level Run Length Matrix 3.6.4 Local Binary Pattern 3.6.5 Gray Level Size Zone Matrix 3.6.6 Local Binary Gray Level Co-Occurrence Matrix 3.6.7 Scale-Invariant Feature Transform 3.6.8 Local Directional Pattern 3.6.9 Segmentation Based Fractal Texture Analysis 3.7 Machine Learning Classifiers 3.7.1 Logistic Regression 3.7.2 Naïve Bayes Classifier 3.7.3 K-Nearest Neighbours 3.7.4 Decision Tree 3.7.5 Support Vector Machine 3.7.6 Random Forest 3.7.7 Bagging 3.7.8 Logit Boost 3.7.9 K-star 3.8 Deep Learning Classifier 3.8.1 Background 3.8.2 Basic Architecture 3.8.2.1 Convolution layer 3.8.2.2 Rectified Linear Unit (ReLU) 3.8.2.3 Pooling Layer 3.8.2.4 Maxpooling 3.8.2.5 Average Pooling 3.8.3 Batch Normalization 3.8.4 Drop-out 3.8.5 Fully Connected Layer 3.8.6 Training Procedure for CNN 3.8.7 Pre-Processing and Augmentation of Data 3.8.8 Initialization of Parameters 3.8.8.1 Random Initialization 3.8.8.2 Unsupervised Pre-Training Initialization 3.8.9 CNN Regularization 3.8.10 Choosing an Optimizer 3.8.10.1 Gradient Descent 3.8.10.2 AdaDelta 3.8.10.3 Adam 3.9 Performance Metrics 3.9.1 Confusion Matrix 3.9.2 Accuracy 3.9.3 Precision 3.9.4 Recall or Sensitivity 3.9.5 Specificity 3.9.6 F1-score 3.10 Pre-Trained Network Models 3.10.1 VGG Architecture 3.10.2 ResNet V2 3.10.3 DenseNet 121 3.10.4 MobileNet 3.11 Conclusion 4. Thermal Imaging for Arthritis evaluation in a Small Animal Model 4.1 Induction and evaluation of arthritis 4.1.1 Collagen Induced Arthritis 4.1.2 Mono arthritis model 4.1.3 Pristane Induced Arthritis 4.1.4 Adjuvant Induced Arthritis 4.1.5 Antigen Induced Arthritis 4.1.6 Proteoglycan (PG) induced ar…