Physics-informed Machine Learning for Medical Image Analysis

A.P. Dhawan, H.K. Huang, D.-S. Kim, eds. World Scientific Publishing; 2008, 868 pages, $228.00. At first glance at the cover and size of Principles and Advanced Methods in Medical Imaging and Image Analysis , a passerby might think that it is the latest Harry Potter novel; however, the gadgets and

Brain development and atrophy accompany people's life. Brain development diseases, such as autism and Alzheimer's disease, affect a large part of the population. Analyzing brain development is very important in public healthcare, and image registration is essential in medical brain image analysis. Many previous studies investigate registration accuracy by the “ground truth” dataset, marker-based similarity calculation, and expert check to find the best registration algorithms. But the evaluation of image registration technology only at the accuracy level is not comprehensive. Here, we compare the performance of three publicly available registration techniques in brain magnetic resonance imaging (MRI) analysis based on some key features widely used in previous MRI studies for classification and detection tasks. According to the analysis results, SPM12 has a stable speed and success rate, and it always works as a guiding tool for newcomers to medical image analysis. It can preserve maximum contrast information, which will facilitate studies such as tumor diagnosis. FSL is a mature and widely applicable toolkit for users, with a relatively stable success rate and good performance. It has complete functions and its function-based integrated toolbox can meet the requirements of different researchers. AFNI is a flexible and complex tool that is more suitable for professional researchers. It retains most details in medical image analysis, which makes it useful in fine-grained analysis such as volume estimation. Our study provides a new idea for comparing registration tools, where tool selection strategy mainly depends on the research task in which the selected tool can leverage its unique advantages.

Whether tracking patient progress for clinical decision making, or investigating novel therapies, automated analysis of Computed Tomography (CT) imaging data is essential for the future of digital radiomics. In digital radiomics, as in all medical imaging, well annotated data is scarce, whereas unlabelled images are relatively plentiful. In other fields of image processing and medical imaging, the application of self-supervised learning (SSL) to large quantities of unlabelled data has resulted in great strides forward for fast, scalable, and interpretable image analysis. In this work we present a new approach applying SSL to CT imaging data which allows for: 1) Improved performance on image classification tasks, based on 2) dramatically reduced quantity of annotated CT imaging data, whilst also 3) enabling easy exploration and interpretation of the image regions. We applied a selection of self-supervised approaches (BYOL, DINO, SimCLR, & inpainting) to CT imaging data. Because the high dimensionality of CT data prevents us from using them directly to out-of the box SSL models, we adopt a 3D patching approach to reduce the dimensionality of the neural net input, and process each patch independently. We train our self-supervised models on public datasets (DeepLesion, NSCLC), and we specialize these models for tumor classification tasks that we evaluate on AstraZeneca sponsored clinical trials. The specialization is done in two different ways: 1) we use the pre-trained SSL model as an encoder that transforms the images of the clinical study into embeddings, on which we apply supervised classification models; 2) we use transfer learning to fine-tune supervised classification models that take the patches directly as inputs. We find that self-supervised pre-training significantly improves the accuracy on tumor classification tasks compared against a supervised learning baseline. Additionally, using the SSL embeddings we build an interactive map of CT imaging data enabling quick and intuitive inspection of the relevant regions. Our findings show that SSL constitutes an important tool for medical imaging analysis. SSL results in models that generalize better, and enable improved downstream interpretability and predictions. Furthermore, well trained SSL models can be re-applied to multiple indications because they are pre-trained on broad and diverse CT imaging data. Citation Format: Leon Fedden, Zhenning Zhang, Khan Baykaner, Qin Li, Lucas Bordeaux. DIME-CT: Self-supervised learning for medical image analysis using patch-based embeddings [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 1937.

Medical image analysis has witnessed remarkable advancements, even surpassing human-level performance in recent years, driven by the rapid development of advanced deep-learning algorithms. However, when the inference dataset slightly differs from what the model has seen during one-time training, the model performance is greatly compromised. The situation requires restarting the training process using both the old and the new data, which is computationally costly, does not align with the human learning process, and imposes storage constraints and privacy concerns. Alternatively, continual learning has emerged as a crucial approach for developing unified and sustainable deep models to deal with new classes, tasks, and the drifting nature of data in non-stationary environments for various application areas. Continual learning techniques enable models to adapt and accumulate knowledge over time, which is essential for maintaining performance on evolving datasets and novel tasks. Owing to its popularity and promising performance, it is an active and emerging research topic in the medical field and hence demands a survey and taxonomy to clarify the current research landscape of continual learning in medical image analysis. This systematic review paper provides a comprehensive overview of the state-of-the-art in continual learning techniques applied to medical image analysis. We present an extensive survey of existing research, covering topics including catastrophic forgetting, data drifts, stability, and plasticity requirements. Further, an in-depth discussion of key components of a continual learning framework, such as continual learning scenarios, techniques, evaluation schemes, and metrics, is provided. Continual learning techniques encompass various categories, including rehearsal, regularization, architectural, and hybrid strategies. We assess the popularity and applicability of continual learning categories in various medical sub-fields like radiology and histopathology. Our exploration considers unique challenges in the medical domain, including costly data annotation, temporal drift, and the crucial need for benchmarking datasets to ensure consistent model evaluation. The paper also addresses current challenges and looks ahead to potential future research directions.

This book places an emphasis on the basic principles and recent advances in medical imaging technologies. After an introduction, the rest of the text can be categorized into three main parts. Part 1 provides a comprehensive tutorial on the fundamentals of medical imaging modalities and image processing methodologies. Part 2 highlights the cutting-edge imaging modalities for radiography and the recent developments of information technology in biomedicine. Part 3 provides some case studies and applications in radiation therapy and computer-aided diagnosis along with a perspective on future trends in medical and molecular imaging. The book is well organized and well presented. A concise summary before each chapter highlights the most important points of the text. This book can be used as a textbook for graduate courses and as an excellent reference for radiologists, neurologists, biomedical engineers, and researchers who are interested in imaging processing and modeling.

Glaucoma is a serious condition of the optic nerve, resulting in loss of eyesight. A sight-threatening disease, it causes impairment of the optic nerve. Glaucoma detection is an important field of medical image analysis. Nowadays, the scope of machine learning in medical image analysis as well as glaucoma detection is drawing attention from the clinical point of view. The ophthalmologist can avail the benefit of one more opinion through computer-aided diagnosis. The algorithm of machine learning can provide accuracy that is one step ahead for detection of glaucoma. Hence from the accuracy point of view, machine learning techniques can be a beneficial experience for the ophthalmologist. Identification of symptoms for glaucoma and glaucoma affected retinal images is a very important task. Machine learning algorithm works on the symptoms and based on the algorithm, it can classify glaucomatous eyes. Machine learning algorithm needs relevant features for identifying the symptom of glaucoma. For extracting accurate features, good quality image is important. Hence, here we mainly focus on the approach used by machine learning techniques for automatic detection of glaucoma. Retinal image acquisition, image preprocessing techniques, feature extraction, classification of symptoms of glaucoma affected image, and evaluating the performance are the important steps in machine learning techniques. Deep learning approach is an also elaborate process for detection of glaucoma in retinal images.

In a variety of medical diagnostic applications, Automatic Defect Detection in clinical imaging has turned into the developing field. Computerized discovery of cancer in MRI which gives the data about the aberrant tissues which is essential for the diagnosis. The traditional technique for Abnormalities detection in Brain is human investigation. This strategy is illogical because of the vast volume of data and the imperfection. Henceforth, trusted and programmed algorithms are preferred to prevent the passing pace of human. In this way, Automated tumor discovery techniques are created as it would save the specialist (radiologist) time and acquire the perfectness. Because of the complexities and diversity of malignancies, MRI brain tumour identification is a difficult task. Machine learning approaches are employed to get over the limitations of traditional classifiers in detecting malignancies in brain scans in this study. MRI scans can be utilised to successfully identify sick cells from healthy ones using machine learning and image classifiers. Convolutional neural network algorithm has been used for classification.

Not-so-supervised: A survey of semi-supervised, multi-instance, and transfer learning in medical image analysis.

Semiautomated methods for microscopic image acquisition, image analysis, and taxonomic identification have repeatedly received attention in diatom analysis. Less well studied is the question whether and how such methods might prove useful for clarifying the delimitation of species that are difficult to separate for human taxonomists. To try to answer this question, three very similar Fragilariopsis species endemic to the Southern Ocean were targeted in this study: F. obliquecostata, F. ritscheri, and F. sublinearis. A set of 501 extended focus depth specimen images were obtained using a standardized, semiautomated microscopic procedure. Twelve diatomists independently identified these specimen images in order to reconcile taxonomic opinions and agree upon a taxonomic gold standard. Using image analyses, we then extracted morphometric features representing taxonomic characters of the target taxa. The discriminating ability of individual morphometric features was tested visually and statistically, and multivariate classification experiments were performed to test the agreement of the quantitatively defined taxa assignments with expert consensus opinion. Beyond an updated differential diagnosis of the studied taxa, our study also shows that automated imaging and image analysis procedures for diatoms are coming close to reaching a broad applicability for routine use.

Dementia, a chronic and progressive cognitive declination of brain function caused by disease or impairment, is becoming more prevalent due to the aging population. A major challenge in dementia is achieving accurate and timely diagnosis. In recent years, neuroimaging with computer-aided algorithms have made remarkable advances in addressing this challenge. The success of these approaches is mostly attributed to the application of machine learning techniques for neuroimaging. In this review paper, we present a comprehensive survey of automated diagnostic approaches for dementia using medical image analysis and machine learning algorithms published in the recent years. Based on the rigorous review of the existing works, we have found that, while most of the studies focused on Alzheimer's disease, recent research has demonstrated reasonable performance in the identification of other types of dementia remains a major challenge. Multimodal imaging analysis deep learning approaches have shown promising results in the diagnosis of these other types of dementia. The main contributions of this review paper are as follows. 1) Based on the detailed analysis of the existing literature, this paper discusses neuroimaging procedures for dementia diagnosis. 2) It systematically explains the most recent machine learning techniques and, in particular, deep learning approaches for early detection of dementia.

Dynamic physiological information recovery : a sampled-data filtering framework by Shan Tong thesis 2008 xiii, 114 leaves : ill. ; 30 cm In medical imaging and image analysis, the goals of a variety of problems, including cardiac motion estimation and…Read more ›

Deep learning models in medical image analysis

Medical image analysis using deep neural networks (DNN) has demonstrated state-of-the-art performance in image classification and segmentation tasks, aiding disease diagnosis. The accuracy of the DNN is largely governed by the quality and quantity of the data used to train the model. However, for the medical images, the critical security and privacy concerns regarding sharing of local medical data across medical establishments precludes exploiting the full DNN potential for clinical diagnosis. The federated learning (FL) approach enables the use of local model's parameters to train a global model, while ensuring data privacy and security. In this paper, we review the federated learning applications in medical image analysis with DNNs, highlight the security concerns, cover some efforts to improve FL model performance, and describe the challenges and future research directions.

With the rise of deep learning, medical image analysis has undergone fundamental change, presenting new possibilities to physicians in terms of achieving higher accuracy more efficiently, and greater automation in disease diagnosis and prognosis. In this paper, we discuss recent progress in utilizing deep learning techniques such as convolutional neural network (CNN), recurrent neural network (RNN), generative adversarial network (GAN), and transformer-based model for medical image analysis. Image classification, segmentation, and anomaly detection in modalities such MRI, CT, Xray, and ultrasound have been strongly improved by these architectures. Yet, such progress brings challenges, namely scarcity of sufficient data, interpretability of models, computational intensity, and robust generalization to diverse clinical datasets. At the same time regulatory concerns and ethical questions on the use of AI diagnostics are still constraining large scale adoption. This paper provides a comprehensive review of state-of-the art deep learning models in medical imaging, reviews the current limitations, and envisions future research directions in federated learning, explainable AI, and quantum enhanced deep learning. Such insights are intended to guide researchers and clinicians in the optimization of AI driven medical image analysis for optimized patient care and clinical decision making.

Deep Learning in Medical Ultrasound Analysis: A Review