Multi-branch CNN and grouping cascade attention for medical image classification

This paper addresses the task of class-imbalanced medical image classification using complication labels. Medical image classification plays an essential role in improving patient care and reducing the burden on healthcare systems, along with other computer aided diagnosis tasks. However, there are two common obstacles with medical image image classification. One is that there is a high cost of acquiring annotations from doctors/experts. Two is that in many cases there is a differences in the number of incidents that appear per symptom which lead to class-imbalance. The above factors make medical image classification difficult to deal with. Additionally, there are cases where multiple symptoms appear in a mixed manner (complications), such as diabetic complications where both symptoms of infection and ischaemia could appear at the same time, which also could influence the classification result negatively. In this paper, a method that jointly tackles the obstacles mentioned above is presented. First, we conduct extensive experiment with various class-imbalanced learning methods introduced in previous works and also propose a method that improves the baseline class-imbalanced approaches by utilizing complication labels in pre-training. Second, we validate the proposed method with the diabetic foot ulcer dataset introduced in Diabetic Foot Ulcer Challenge 2021.

Medical Imaging & Informatics has become a fast‐growing market field in the medical private industry. It is driven mostly by health care institutions' desire to provide better quality service in decreasing turnaround time of diagnosis. In addition, Imaging and Informatics is moving from a purely Radiology product to an Enterprise‐wide product. With these forces driving the market forward, there is an apparent lack of resources both in the private industry as well as in health care institutions to properly address each of their needs. Because the clinical environment is a high‐demand, 24/7 operation, it presents a unique set of factors that other Information Technology (IT)‐related fields may not encounter. In addition, the image data that is involved can be very large in size in comparison to conventional, mostly text‐based IT‐related data. Traditionally, private industries have hired personnel with an IT background to lead the medical imaging & informatics product roadmap development and implementation. However, most people in the IT field lack the experience or knowledge of the clinical environment where the products will be embedded. This is crucial for the success of both the health care institution and the vendor. As for a health care institution, most personnel hired to manage, support, and implement a medical imaging & informatics product (eg, Picture Archive Communication System, or PACS) have a clinical background with limited experience in large‐scale IT system design and implementation which again, is crucial for success. The Medical ImagingInformatics field bridges the gap between private industry and the health care institution. It leverages the knowledge and expertise of IT‐related skills (eg, large‐scale system design, system uptime, acceptance testing, disaster recovery, etc) with the experience of how these medical imaging and informatics applications are utilized within the clinical environment. Because there is a constant flow of research within this unique relationship with the clinical environment, medical imaginginformatics also drives research and development of new products to remain on the leading edge of technology as well as determining the current market demands and trends. With these strengths, the Biomedical Engineering Master's Program in Medical Imaging and ImagingInformatics, Viterbi School of Engineering, University of Southern California draws course attendees from medical industry corporations already heavily involved in medical imaging and informatics products. Personnel from other IT‐related corporations looking for a career change into the medical arena are potential attendees as well. Clinical staff in a health care institution who are interested in branching out of their own respective clinical environments into the new arena of designing, implementing, and managing medical imaging and informatics systems such as PACS would also be ideal course attendees. One of the challenges is the uniqueness of this particular field of Medical ImagingInformatics is the lack of courses offered across the country. Therefore, the two courses offered in Medical ImagingInformatics are also taught through the Distance Education Network for students who are offsite from the USC campus. The following presentation will discuss course content, tools, and challenges/pitfalls that accompany distance education and learning specifically in the area of Medical ImagingInformatics. The following are learning objectives for the following presentation: 1. Learn about distance education network at USC 2. Learn about course content for Medical ImagingInformatics 3. Challenges in disseminating course content and student feedback 4. Learn what tools are helpful to facilitate distance learning for medical ImagingInformatics 5. See a demonstration of what a DEN classroom looks like from the off‐campus student perspective

In recent years, generative adversarial networks (GANs) have gained tremendous popularity for various imaging related tasks such as artificial image generation to support AI training. GANs are especially useful for medical imaging-related tasks where training datasets are usually limited in size and heavily imbalanced against the diseased class. We present a systematic review, following the PRISMA guidelines, of recent GAN architectures used for medical image analysis to help the readers in making an informed decision before employing GANs in developing medical image classification and segmentation models. We have extracted 54 papers that highlight the capabilities and application of GANs in medical imaging from January 2015 to August 2020 and inclusion criteria for meta-analysis. Our results show four main architectures of GAN that are used for segmentation or classification in medical imaging. We provide a comprehensive overview of recent trends in the application of GANs in clinical diagnosis through medical image segmentation and classification and ultimately share experiences for task-based GAN implementations.

Due to the complexity of medical images, traditional medical image classification methods have been unable to meet the actual application needs. In recent years, the rapid development of deep learning theory has provided a technical approach for solving medical image classification. However, deep learning has the following problems in the application of medical image classification. First, it is impossible to construct a deep learning model with excellent performance according to the characteristics of medical images. Second, the current deep learning network structure and training strategies are less adaptable to medical images. Therefore, this paper first introduces the visual attention mechanism into the deep learning model so that the information can be extracted more effectively according to the problem of medical images, and the reasoning is realized at a finer granularity. It can increase the interpretability of the model. Additionally, to solve the problem of matching the deep learning network structure and training strategy to medical images, this paper will construct a novel multiscale convolutional neural network model that can automatically extract high-level discriminative appearance features from the original image, and the loss function uses the Mahalanobis distance optimization model to obtain a better training strategy, which can improve the robust performance of the network model. The medical image classification task is completed by the above method. Based on the above ideas, this paper proposes a medical classification algorithm based on a visual attention mechanism-multiscale convolutional neural network. The lung nodules and breast cancer images were classified by the method in this paper. The experimental results show that the accuracy of medical image classification in this paper is not only higher than that of traditional machine learning methods but also improved compared with other deep learning methods, and the method has good stability and robustness.

Image patch classification is an important task in many different medical imaging applications. In this work, we have designed a customized Convolutional Neural Networks (CNN) with shallow convolution layer to classify lung image patches with interstitial lung disease (ILD). While many feature descriptors have been proposed over the past years, they can be quite complicated and domain-specific. Our customized CNN framework can, on the other hand, automatically and efficiently learn the intrinsic image features from lung image patches that are most suitable for the classification purpose. The same architecture can be generalized to perform other medical image or texture classification tasks.

Bridging 2D and 3D segmentation networks for computation-efficient volumetric medical image segmentation: An empirical study of 2.5D solutions.

Deep learning methods, especially convolutional neural networks, have become more and more popular in medical image classifications. However, training a deep neural network from scratch can be a luxury for many medical image datasets as the process requires a large and well-balanced sample to output satisfactory results. Unlike natural image datasets, medical images are expensive to collect owing to labor and equipment costs. Besides, the class labels in medical image datasets are usually severely imbalanced subject to the availability of patients. Further, aggregating medical images from multiple sources can be challenging due to policy restrictions, privacy concerns, communication costs, and data heterogeneity caused by equipment differences and labeling discrepancies. In this paper, we propose to address these issues with the help of transfer learning and artificial samples created by generative models. Instead of requesting medical images from source data, our method only needs a parsimonious supplement of model parameters pre-trained on the source data. The proposed method preserves the data privacy in the source data and significantly reduces the communication cost. Our study shows transfer learning together with artificial samples can improve the pneumonia classification accuracy on a small but heavily imbalanced chest X-ray image dataset by \(11.53\%\) which performs even better than directly augmenting that source data into the training process.KeywordsDeep learningGenerative modelsMedical image classificationPrivacy preservationTransfer learning

Due to the complexity of medical images, traditional medical image classification methods have been unable to meet actual application needs. In recent years, the rapid development of deep learning theory has provided a technical approach for solving medical image classification tasks. However, deep learning has the following problems in medical image classification. First, it is impossible to construct a deep learning model hierarchy for medical image properties; second, the network initialization weights of deep learning models are not well optimized. Therefore, this paper starts from the perspective of network optimization and improves the nonlinear modeling ability of the network through optimization methods. A new network weight initialization method is proposed, which alleviates the problem that existing deep learning model initialization is limited by the type of the nonlinear unit adopted and increases the potential of the neural network to handle different visual tasks. Moreover, through an in‐depth study of the multicolumn convolutional neural network framework, this paper finds that the number of features and the convolution kernel size at different levels of the convolutional neural network are different. In contrast, the proposed method can construct different convolutional neural network models that adapt better to the characteristics of the medical images of interest and thus can better train the resulting heterogeneous multicolumn convolutional neural networks. Finally, using the adaptive sliding window fusion mechanism proposed in this paper, both methods jointly complete the classification task of medical images. Based on the above ideas, this paper proposes a medical classification algorithm based on a weight initialization/sliding window fusion for multilevel convolutional neural networks. The methods proposed in this study were applied to breast mass, brain tumor tissue, and medical image database classification experiments. The results show that the proposed method not only achieves a higher average accuracy than that of traditional machine learning and other deep learning methods but also is more stable and more robust.

Medical supervised masked autoencoder: Crafting a better masking strategy and efficient fine-tuning schedule for medical image classification

We previously developed two AI-based medical automatic image classification tools using a multi-layer fuzzy approach (MFA and MCM) to convert image-based abnormality into a quantity. However, there is currently limited research on using diagnostic image assessment tools to statistically predict the hazard due to the disease. The present study introduces a novel approach that addresses a substantial research gap in the identification of hazard or risk associated with a disease using an automatically quantified image-based abnormality. The method employed to ascertain hazard in an image-based quantified abnormality was the cox proportional hazard (PH) model, a unique tool in medical research for identifying hazard related to covariates. MFA was first used to quantify the abnormality in CT scan images, and hazard plots were utilized to visually represent the hazard risk over time. Hazards corresponding to image-based abnormality were then computed for the variables, ‘gender,’ ‘age,’ and ‘smoking-status’. This integrated framework potentially minimizes false negatives, identifies patients with the highest mortality risk and facilitates timely initiation of treatment. By utilizing pre-existing patient images, this method could reduce the considerable costs associated with public health research and clinical trials. Furthermore, understanding the hazard posed by widespread global diseases like COVID-19 aids medical researchers in prompt decision-making regarding treatment and preventive measures.

For the transmission, storage or telemedicine applications patient information is embedded inside the medical images. The purpose of data hiding in medical image is to embed the patient information in the host medical image without perceptually distorting it. This paper proposes a novel approach in embedding Patient information and Electrocardiogram (Patient Data) which is encrypted to ensure greater security in spatial domain. Statistical parameters such as Normalized Root Mean Square Error and Peak Signal to Noise Ratio are used to measure the quality of the medical image with patient data. Experimental results strongly indicate that the present technique is found to be robust and embedded patient data is recoverable without any distortion even in the case of attack like Gaussian noise.

Label noise in medical image classification datasets significantly hampers the training of supervised deep learning methods, undermining their generalizability. The test performance of a model tends to decrease as the label noise rate increases. Over recent years, several methods have been proposed to mitigate the impact of label noise in medical image classification and enhance the robustness of the model. Predominantly, these works have employed CNN-based architectures as the backbone of their classifiers for feature extraction. However, in recent years, Vision Transformer (ViT)-based backbones have replaced CNNs, demonstrating improved performance and a greater ability to learn more generalizable features, especially when the dataset is large. Nevertheless, no prior work has rigorously investigated how transformer-based backbones handle the impact of label noise in medical image classification. In this paper, we investigate the architectural robustness of ViT against label noise and compare it to that of CNNs. We use two medical image classification datasets-COVID-DU-Ex, and NCT-CRC-HE-100K-both corrupted by injecting label noise at various rates. Additionally, we show that pretraining is crucial for ensuring ViT's improved robustness against label noise in supervised training.

TC-Net: A joint learning framework based on CNN and vision transformer for multi-lesion medical images segmentation

Content-based medical image classification using a new hierarchical merging scheme

Probabilistic smooth attention for deep multiple instance learning in medical imaging