Improved Deep Semantic Medical Image Segmentation PDF Download
Are you looking for read ebook online? Search for your book and save it on your Kindle device, PC, phones or tablets. Download Improved Deep Semantic Medical Image Segmentation PDF full book. Access full book title Improved Deep Semantic Medical Image Segmentation by Saeid Asgari Taghanaki. Download full books in PDF and EPUB format.
Author: Saeid Asgari Taghanaki Publisher: ISBN: Category : Languages : en Pages : 132
Book Description
The image semantic segmentation challenge consists of classifying each pixel of an image (or just several ones) into an instance, where each instance (or category) corresponds to an object. This task is a part of the concept of scene understanding or better explaining the global context of an image. In the medical image analysis domain, image segmentation can be used for image-guided interventions, radiotherapy, or improved radiological diagnostics. Following a comprehensive review of state-of-the-art deep learning-based medical and non-medical image segmentation solutions, we make the following contributions. A deep learning-based (medical) image segmentation typical pipeline includes designing layers (A), designing an architecture (B), and defining a loss function (C). A clean/modified (D)/adversarialy perturbed (E) image is fed into a model (consisting of layers and loss function) to predict a segmentation mask for scene understanding etc. In some cases where the number of segmentation annotations is limited, weakly supervised approaches (F) are leverages. For some applications where further analysis is needed e.g., predicting volumes and objects burden, the segmentation mask is fed into another post-processing step (G). In this thesis, we tackle each of the steps (A-G). I) As for step (A and E), we studied the effect of the adversarial perturbation on image segmentation models and proposed a method that improves the segmentation performance via a non-linear radial basis convolutional feature mapping by learning a Mahalanobis-like distance function on both adversarially perturbed and unperturbed images. Our method then maps the convolutional features onto a linearly well-separated manifold, which prevents small adversarial perturbations from forcing a sample to cross the decision boundary. II) As for step (B), we propose light, learnable skip connections which learn first to select the most discriminative channels and then aggregate the selected ones as single-channel attending to the most discriminative regions of input. Compared to the heavy classical skip connections, our method reduces the computation cost and memory usage while it improves segmentation performance. III) As for step (C), we examined the critical choice of a loss function in order to handle the notorious imbalance problem that plagues both the input and output of a learning model. In order to tackle both types of imbalance during training and inference, we introduce a new curriculum learning-based loss function. Specifically, we leverage the Dice similarity coefficient to deter model parameters from being held at bad local minima and at the same time, gradually learn better model parameters by penalizing for false positives/negatives using a cross-entropy term which also helps. IV) As for step (D), we propose a new segmentation performance-boosting paradigm that relies on optimally modifying the network's input instead of the network itself. In particular, we leverage the gradients of a trained segmentation network with respect to the input to transfer it into a space where the segmentation accuracy improves. V) As for step (F), we propose a weakly supervised image segmentation model with a learned spatial masking mechanism to filter out irrelevant background signals from attention maps. The proposed method minimizes mutual information between a masked variational representation and the input while maximizing the information between the masked representation and class labels. VI) Although many semi-automatic segmentation based methods have been developed, as for step (G), we introduce a method that completely eliminates the segmentation step and directly estimates the volume and activity of the lesions from positron emission tomography scans.
Author: Saeid Asgari Taghanaki Publisher: ISBN: Category : Languages : en Pages : 132
Book Description
The image semantic segmentation challenge consists of classifying each pixel of an image (or just several ones) into an instance, where each instance (or category) corresponds to an object. This task is a part of the concept of scene understanding or better explaining the global context of an image. In the medical image analysis domain, image segmentation can be used for image-guided interventions, radiotherapy, or improved radiological diagnostics. Following a comprehensive review of state-of-the-art deep learning-based medical and non-medical image segmentation solutions, we make the following contributions. A deep learning-based (medical) image segmentation typical pipeline includes designing layers (A), designing an architecture (B), and defining a loss function (C). A clean/modified (D)/adversarialy perturbed (E) image is fed into a model (consisting of layers and loss function) to predict a segmentation mask for scene understanding etc. In some cases where the number of segmentation annotations is limited, weakly supervised approaches (F) are leverages. For some applications where further analysis is needed e.g., predicting volumes and objects burden, the segmentation mask is fed into another post-processing step (G). In this thesis, we tackle each of the steps (A-G). I) As for step (A and E), we studied the effect of the adversarial perturbation on image segmentation models and proposed a method that improves the segmentation performance via a non-linear radial basis convolutional feature mapping by learning a Mahalanobis-like distance function on both adversarially perturbed and unperturbed images. Our method then maps the convolutional features onto a linearly well-separated manifold, which prevents small adversarial perturbations from forcing a sample to cross the decision boundary. II) As for step (B), we propose light, learnable skip connections which learn first to select the most discriminative channels and then aggregate the selected ones as single-channel attending to the most discriminative regions of input. Compared to the heavy classical skip connections, our method reduces the computation cost and memory usage while it improves segmentation performance. III) As for step (C), we examined the critical choice of a loss function in order to handle the notorious imbalance problem that plagues both the input and output of a learning model. In order to tackle both types of imbalance during training and inference, we introduce a new curriculum learning-based loss function. Specifically, we leverage the Dice similarity coefficient to deter model parameters from being held at bad local minima and at the same time, gradually learn better model parameters by penalizing for false positives/negatives using a cross-entropy term which also helps. IV) As for step (D), we propose a new segmentation performance-boosting paradigm that relies on optimally modifying the network's input instead of the network itself. In particular, we leverage the gradients of a trained segmentation network with respect to the input to transfer it into a space where the segmentation accuracy improves. V) As for step (F), we propose a weakly supervised image segmentation model with a learned spatial masking mechanism to filter out irrelevant background signals from attention maps. The proposed method minimizes mutual information between a masked variational representation and the input while maximizing the information between the masked representation and class labels. VI) Although many semi-automatic segmentation based methods have been developed, as for step (G), we introduce a method that completely eliminates the segmentation step and directly estimates the volume and activity of the lesions from positron emission tomography scans.
Author: Le Lu Publisher: Springer ISBN: 331942999X Category : Computers Languages : en Pages : 327
Book Description
This book presents a detailed review of the state of the art in deep learning approaches for semantic object detection and segmentation in medical image computing, and large-scale radiology database mining. A particular focus is placed on the application of convolutional neural networks, with the theory supported by practical examples. Features: highlights how the use of deep neural networks can address new questions and protocols, as well as improve upon existing challenges in medical image computing; discusses the insightful research experience of Dr. Ronald M. Summers; presents a comprehensive review of the latest research and literature; describes a range of different methods that make use of deep learning for object or landmark detection tasks in 2D and 3D medical imaging; examines a varied selection of techniques for semantic segmentation using deep learning principles in medical imaging; introduces a novel approach to interleaved text and image deep mining on a large-scale radiology image database.
Author: Le Lu Publisher: Springer Nature ISBN: 3030139697 Category : Computers Languages : en Pages : 461
Book Description
This book reviews the state of the art in deep learning approaches to high-performance robust disease detection, robust and accurate organ segmentation in medical image computing (radiological and pathological imaging modalities), and the construction and mining of large-scale radiology databases. It particularly focuses on the application of convolutional neural networks, and on recurrent neural networks like LSTM, using numerous practical examples to complement the theory. The book’s chief features are as follows: It highlights how deep neural networks can be used to address new questions and protocols, and to tackle current challenges in medical image computing; presents a comprehensive review of the latest research and literature; and describes a range of different methods that employ deep learning for object or landmark detection tasks in 2D and 3D medical imaging. In addition, the book examines a broad selection of techniques for semantic segmentation using deep learning principles in medical imaging; introduces a novel approach to text and image deep embedding for a large-scale chest x-ray image database; and discusses how deep learning relational graphs can be used to organize a sizable collection of radiology findings from real clinical practice, allowing semantic similarity-based retrieval. The intended reader of this edited book is a professional engineer, scientist or a graduate student who is able to comprehend general concepts of image processing, computer vision and medical image analysis. They can apply computer science and mathematical principles into problem solving practices. It may be necessary to have a certain level of familiarity with a number of more advanced subjects: image formation and enhancement, image understanding, visual recognition in medical applications, statistical learning, deep neural networks, structured prediction and image segmentation.
Author: S. Kevin Zhou Publisher: Academic Press ISBN: 0323858880 Category : Computers Languages : en Pages : 544
Book Description
Deep Learning for Medical Image Analysis, Second Edition is a great learning resource for academic and industry researchers and graduate students taking courses on machine learning and deep learning for computer vision and medical image computing and analysis. Deep learning provides exciting solutions for medical image analysis problems and is a key method for future applications. This book gives a clear understanding of the principles and methods of neural network and deep learning concepts, showing how the algorithms that integrate deep learning as a core component are applied to medical image detection, segmentation, registration, and computer-aided analysis. · Covers common research problems in medical image analysis and their challenges · Describes the latest deep learning methods and the theories behind approaches for medical image analysis · Teaches how algorithms are applied to a broad range of application areas including cardiac, neural and functional, colonoscopy, OCTA applications and model assessment · Includes a Foreword written by Nicholas Ayache
Author: Ben Othman Soufiene Publisher: CRC Press ISBN: 1003805671 Category : Technology & Engineering Languages : en Pages : 270
Book Description
Machine Learning and Deep Learning Techniques for Medical Image Recognition comprehensively reviews deep learning-based algorithms in medical image analysis problems including medical image processing. It includes a detailed review of deep learning approaches for semantic object detection and segmentation in medical image computing and large-scale radiology database mining. A particular focus is placed on the application of convolutional neural networks with the theory and varied selection of techniques for semantic segmentation using deep learning principles in medical imaging supported by practical examples. Features: Offers important key aspects in the development and implementation of machine learning and deep learning approaches toward developing prediction tools and models and improving medical diagnosis Teaches how machine learning and deep learning algorithms are applied to a broad range of application areas, including chest X-ray, breast computer-aided detection, lung and chest, microscopy, and pathology Covers common research problems in medical image analysis and their challenges Focuses on aspects of deep learning and machine learning for combating COVID-19 Includes pertinent case studies This book is aimed at researchers and graduate students in computer engineering, artificial intelligence and machine learning, and biomedical imaging.
Author: Darvin Yi Publisher: ISBN: Category : Languages : en Pages :
Book Description
The rise of deep learning (DL) has created many novel algorithms for segmentation, which has in turn revolutionized the field of medical image segmentation. However, several distinctions between the field of natural and medical computer vision necessitates specialized algorithms to optimize performance, including the multi-modality of medical data, the differences in imaging protocols between centers, and the limited amount of annotated data. These differences lead to limitations when applying current state of the art computer vision methods on medical imaging. For segmentation, the major gaps our algorithms must bridge to become clinically useful are: (1) generalize to different imaging protocols, (2) become robust to training on noisy labels, and (3) generally improve segmentation performance. The current rigorous deep learning architectures are not robust to having missing input modalities after training a network, which makes our networks unable to run inference on new data taken with a different imaging protocol. By training our algorithms without taking into account the mutability of imaging protocols, we heavily limit the deployability of our algorithms. Our current training paradigm also needs pristine segmentation labels, which necessitates a large time investment by expert annotators. By training our algorithms with an underlying assumption that there is no noise in our labels with harsh loss functions like cross entropy, we create a need for clean labels. This limits our datasets from being fully largely scalable to the same size as natural computer vision datasets, as disease segmentations on medical images require more time and effort to annotate than natural images with semantic classes. Finally, current state of the art performance on difficult segmentation tasks like brain metastases is just not enough to be clinically useful. We will need to explore new ways of designing and ensembling networks to increase segmentation performance should we aim to deploy these algorithms in any clinically relevant environment. We hypothesize that by changing neural network architectures and loss functions to account for noisy data rather than assuming consistent imaging protocols and pristine labels, we can encode more robustness into our trained networks and improve segmentation performance on medical imaging tasks. In our experiments, we will test several different networks whose architecture and loss functions have been motivated by realistic and clinically relevant situations. For these experiments, we chose the model system of brain metastases lesion detection and segmentation, a difficult problem due to the high count and small size of the lesions. It is also an important problem due to the need to assess the effects of treatment by tracking changes in tumor burden. In this dissertation, we present the following specific aims: (1) optimizing deep learning performance on brain metastases segmentation, (2) training networks to be robust to coarse annotations and missing data, and (3) validating our methodology on three different secondary tasks. Our trained baseline performance (state of the art) performs brain metastases segmentation modestly, giving us mAP values of $0.46\pm0.02$ and DICE scores of 0.72. Changing our architectures to account for different pulse sequence integration methods does not improve our values by much, giving us a model mAP improvement to $0.48\pm0.2$ and no improvement in DICE score. However, through investigating pulse sequence integration, we developed a novel input-level dropout training scheme that holds out certain pulse sequences randomly during different iterations of training our deep net. This trains our network to be robust to missing pulse sequences in the future, at no cost to performance. We then developed two additional robustness training schemes that enable training on data annotations that have a lot of noise. We prove that we are able to lose no performance when degrading 70\% of our segmentation annotations with spherical approximations, and show a loss of 5\% performance when degrading 90\% of our annotations. Similarly, when we censor our 50\% of our annotated lesions (simulating a 50\% False Negative Rate), we can preserve 95\% of the performance by utilizing a novel lopsided bootstrap loss. Using these ideas, we use the lesion-based censoring technique as the base of a novel ensembling method we named Random Bundle. This network increased our mAP value $0.65\pm0.01$, an increase of about 40\%. We validate our methods on three different secondary datasets. By validating our methods work on brain metastases data from Oslo University Hospital, we show that our methods are robust to cross-center data. By validating our methods on the MICCAI BraTS dataset, we show that our methods are robust to magnetic resonance images of a different disorder. Finally, by validating our methods on diabetic retinopathy micro-aneurysms on fundus photographs, we show that our methods are robust across imaging domains and organ systems. Our experiments support our claims that (1) designing architectures with a focus on how pulse sequences interact will encode robustness for different imaging protocols, (2) creating custom loss functions around expected annotation errors will make our networks more robust to those errors, and (3) the overall performance of our networks can be improved by using these novel architectures and loss functions.
Author: Saxena, Sanjay Publisher: IGI Global ISBN: 1799850722 Category : Medical Languages : en Pages : 274
Book Description
Before the modern age of medicine, the chance of surviving a terminal disease such as cancer was minimal at best. After embracing the age of computer-aided medical analysis technologies, however, detecting and preventing individuals from contracting a variety of life-threatening diseases has led to a greater survival percentage and increased the development of algorithmic technologies in healthcare. Deep Learning Applications in Medical Imaging is a pivotal reference source that provides vital research on the application of generating pictorial depictions of the interior of a body for medical intervention and clinical analysis. While highlighting topics such as artificial neural networks, disease prediction, and healthcare analysis, this publication explores image acquisition and pattern recognition as well as the methods of treatment and care. This book is ideally designed for diagnosticians, medical imaging specialists, healthcare professionals, physicians, medical researchers, academicians, and students.
Author: Alejandro Frangi Publisher: Academic Press ISBN: 0128136588 Category : Technology & Engineering Languages : en Pages : 700
Book Description
Medical Image Analysis presents practical knowledge on medical image computing and analysis as written by top educators and experts. This text is a modern, practical, self-contained reference that conveys a mix of fundamental methodological concepts within different medical domains. Sections cover core representations and properties of digital images and image enhancement techniques, advanced image computing methods (including segmentation, registration, motion and shape analysis), machine learning, how medical image computing (MIC) is used in clinical and medical research, and how to identify alternative strategies and employ software tools to solve typical problems in MIC. Provides an authoritative description of key concepts and methods Includes tutorial-based sections that clearly explain principles and their application to different medical domains Presents a representative selection of topics to match a modern and relevant approach to medical image computing
Author: Danail Stoyanov Publisher: Springer ISBN: 3030008894 Category : Computers Languages : en Pages : 401
Book Description
This book constitutes the refereed joint proceedings of the 4th International Workshop on Deep Learning in Medical Image Analysis, DLMIA 2018, and the 8th International Workshop on Multimodal Learning for Clinical Decision Support, ML-CDS 2018, held in conjunction with the 21st International Conference on Medical Imaging and Computer-Assisted Intervention, MICCAI 2018, in Granada, Spain, in September 2018. The 39 full papers presented at DLMIA 2018 and the 4 full papers presented at ML-CDS 2018 were carefully reviewed and selected from 85 submissions to DLMIA and 6 submissions to ML-CDS. The DLMIA papers focus on the design and use of deep learning methods in medical imaging. The ML-CDS papers discuss new techniques of multimodal mining/retrieval and their use in clinical decision support.
Author: Tao Lei Publisher: John Wiley & Sons ISBN: 111985900X Category : Technology & Engineering Languages : en Pages : 340
Book Description
Image Segmentation Summarizes and improves new theory, methods, and applications of current image segmentation approaches, written by leaders in the field The process of image segmentation divides an image into different regions based on the characteristics of pixels, resulting in a simplified image that can be more efficiently analyzed. Image segmentation has wide applications in numerous fields ranging from industry detection and bio-medicine to intelligent transportation and architecture. Image Segmentation: Principles, Techniques, and Applications is an up-to-date collection of recent techniques and methods devoted to the field of computer vision. Covering fundamental concepts, new theories and approaches, and a variety of practical applications including medical imaging, remote sensing, fuzzy clustering, and watershed transform. In-depth chapters present innovative methods developed by the authors—such as convolutional neural networks, graph convolutional networks, deformable convolution, and model compression—to assist graduate students and researchers apply and improve image segmentation in their work. Describes basic principles of image segmentation and related mathematical methods such as clustering, neural networks, and mathematical morphology. Introduces new methods for achieving rapid and accurate image segmentation based on classic image processing and machine learning theory. Presents techniques for improved convolutional neural networks for scene segmentation, object recognition, and change detection, etc. Highlights the effect of image segmentation in various application scenarios such as traffic image analysis, medical image analysis, remote sensing applications, and material analysis, etc. Image Segmentation: Principles, Techniques, and Applications is an essential resource for undergraduate and graduate courses such as image and video processing, computer vision, and digital signal processing, as well as researchers working in computer vision and image analysis looking to improve their techniques and methods.
Author: Saeid Asgari Taghanaki
Author: Saeid Asgari Taghanaki
Author: Le Lu
Author: Le Lu
Author: S. Kevin Zhou
Author: Ben Othman Soufiene
Author: Darvin Yi
Author: Saxena, Sanjay
Author: Alejandro Frangi
Author: Danail Stoyanov
Author: Tao Lei