Search

The image segmentation of diseases can help clinical diagnosis and treatment in medical image analysis. Due to the complexity of lesion features (e.g., size, location, and morphology) and the high similarity between the background and the target area in medical images, semantic features are difficult to extract completely. To tackle these problems, we propose a novel medical image segmentation network PAMSNet based on the spatial pyramid and attention mechanism. By using efficient pyramid attention and channel spatial attention modules, the proposed method fuses the extracted multi-scale spatial information with the local features extracted by the encoder to supplement the image details. In addition, the Spatial Pyramid-Coordinate Attention (SPCA) module is introduced in the bottleneck layer to obtain larger receptive field information and enhance feature extraction. We conducted qualitative and quantitative evaluations on four public datasets, including ISIC2018, Lung segmentation, Kvasir-SEG, and ISLES2022. The segmentation accuracy of DSC was 87.86%, 98.18%, 82.43%, and 87.37%, respectively. The ablation study of each part of PAMSNet proves the validity of each component, and the comparison with state-of-the-art methods on different indicators proves the predominance of the network. • PAEMNet extracts detailed spatial info effectively in encoding. • It emphasizes edge and detail features for better segmentation. • The model improves performance significantly on lesion segmentation. [ABSTRACT FROM AUTHOR]

Using deep learning to improve the performance of finger vein recognition has become a part of mainstream research. Currently, the performance of finger vein recognition systems is limited by insufficient finger vein training samples, which leads to insufficient feature learning and weak model generalization. To solve the above problems, first, we propose a simple and effective finger vein data augmentation strategy named FV-Mix, which uses fine finger vein image region of interest (ROI) for grayscale normalization and linear mixing. It can accomplish exponential-level data augmentation on the training sample, and the augmented data will represent a more complete dataset. Second, a residual Gabor convolutional network (RGCN) is designed for finger vein recognition, which includes a residual Gabor convolutional layer (RGCL) and a dense semantic analysis module (DSAM). The RGCL is designed to replace the shallow convolutional layer in the network by using the characteristics of the Gabor filter to enhance the scale and direction of information in the shallow pattern features. And then, the enhanced deep features are further extracted and analyzed by DSAM, which is used to assist the final model in classifying and recognizing finger vein images. To verify the effectiveness of our work, five publicly available finger vein image datasets are used, and a large number of comparison experiments are designed. The proposed FV-Mix strategy, RGCL module, and DSAM module were fully validated. The experimental results show that on these five datasets, the average recognition accuracy and equal error rate (EER) of the proposed RGCN are 99.22% and 0.188%, respectively, achieves competitive performance compared with current state-of-the-art work. • Novel residual Gabor convolutional network improves recognition system performance. • Novel residual Gabor convolutional layer enhances finger vein pattern features. • Novel dense semantic analysis module assists with model classification. • Exponential level augmentation for finger vein image samples. [ABSTRACT FROM AUTHOR]

Deep-learning-based segmentation methods have shown great success across many medical image applications. However, the custom training paradigms suffer from a well-known constraint of the requirement of pixel-wise annotations, which is labor-intensive, especially when they are required to learn new classes incrementally. Contemporary incremental learning focuses on dealing with catastrophic forgetting in image classification and object detection. However, this work aims to promote the performance of the current model to learn new classes with the help of the previous model in the context of incremental learning of instance segmentation. It enormously benefits the current model when the labeled data is limited because of the high labor intensity of manual labeling. In this paper, on the Diabetic Retinopathy (DR) lesion segmentation problem, a novel incremental segmentation paradigm is proposed to distill the knowledge of the previous model to improve the current model. Remarkably, we propose various approaches working on the class-based alignment of the probability maps of the current and the previous model, accounting for the difference between the background classes of the two models. The experimental evaluation of DR lesion segmentation shows the effectiveness of the proposed approaches. • It proposed a scheme for incremental segmentation. • Using knowledge distillation to transfer the knowledge from model to model. • Improve the segmentation performance using incremental learning. • Proposed a probability-map alignment scheme to integrate two different class maps. • The proposed method is generalized to any new classes. [ABSTRACT FROM AUTHOR]

Automatic medical image segmentation method is highly needed to help experts in lesion segmentation. The deep learning technology emerging has profoundly driven the development of medical image segmentation. While U-Net and attention mechanisms are widely utilized in this field, the application of attention, albeit successful in natural scene image segmentation, tends to inflate the number of model parameters and neglects the potential for feature fusion between different convolutional layers. In response to these challenges, we present the Multi-Attention and Depthwise Separable Convolution U-Net (MDSU-Net), designed to enhance feature extraction. The multi-attention aspect of our framework integrates dual attention and attention gates, adeptly capturing rich contextual details and seamlessly fusing features across diverse convolutional layers. Additionally, our encoder integrates a depthwise separable convolution layer, streamlining the model's complexity without sacrificing its efficacy, ensuring versatility across various segmentation tasks. The results demonstrate that our method outperforms state-of-the-art across three diverse medical image datasets. [ABSTRACT FROM AUTHOR]