ESA-Net: An Efficient and Lightweight Model for Medical Image Segmentation

Shape and boundary-aware multi-branch model for semi-supervised medical image segmentation

Taming large vision model for medical image segmentation via Dual Visual Prompt Tuning.

An improved hybrid model (FCM_MS) for medical image segmentation is proposed by combining fuzzy C-means (FCM) clustering and Mumford-Shah (MS) algorithm. In the proposed model, fuzzy membership degree from FCM clustering is firstly used to initialize the contour placement, and then incorporated into the fidelity term of the 2-phase piecewise constant MS model to obtain multi-object segmentation. Meanwhile penalizing energy term is introduced into the energy functional to eliminate re-initialization of level set and thus to fasten convergent speed on curve evolution. Experimental results show that the proposed model has advantages both in accuracy and in robustness to noise in comparison with the standard FCM or the classical MS model on medical image segmentation.

In this work, a Multi-Layer Deformable Model (MLDM) for medical image segmentation is proposed. In contrast to common deformable model based segmentation approaches our new method incorporates a multi-layer geometric model that allows a sampling of the organ's interior. An adaptation logic processes the additional information gained from interior layers in order to fit the model to the data. The deformation is coupled with a dynamic internal energy function represented by a link-oriented flexibility in order to allow the model to accurately adapt to cavities. Exploiting the additional depth information, our approach detects low contrasted transitions between organs more reliably and recovers better from bad model initialization than existing methods. Our approach has been evaluated using representative CT data sets of the liver as well as CT bladder scans. Evaluation using ground truth data showed that our multi-layer technique yields superior results in contrast to common single surface segmentation. Since the amount of layers is flexible, the most interior regions which only carry little regional information can be excluded from optimization. Together with the linear nature of MLDM optimization our approach outperforms other volumetric segmentation methods in terms of speed.

This study proposes an innovative M-L (Multiple-channel local binary fitting) model for medical image segmentation. Designed to improve upon existing image segmentation models, the M-L model introduces a regional limit function to the multi-band active contour model to enable multilayer image segmentation. The Gaussian kernel function is used to improve the previous model's robustness, necessitating the use of a new initialization curve which enhances the accuracy of segmentation results. Compared to existing image segmentation methods, the proposed M-L model improves numerical stability and efficiency through the introduction of a new penalty term and an increased step length. This simulation experiment verifies the advantages of the new M-L model for improved medical image segmentation, including increased efficiency and usability of the model.

To address the loss problem introduced by downsampling in the classical U-Net architecture, this paper improves the U-Net model and uses the model for medical image segmentation. The essence of the proposed model is still the classical U-Net encoder-decoder network, in which the encoder and decoder sub-networks are connected by a skip connection. First, we improve the connection position of the skip connection, and the two ends of the connection are changed from the second convolution result of the original convolution block to the first result and the decoder convolution block for concatenation; second, the concatenation operation is added to the convolution block of the downsampling part, and the two improvements aim at retaining more image underlying information, and thus achieving more efficient fusion of high and low level image information; finally, on the public medical image segmentation dataset, the classical U-Net, FCN-8s and the improved model in this paper are comparatively evaluated for cell nucleus segmentation in microscope images and liver segmentation in abdominal CT scans. The experiments show that the improved U-Net model mIoU, Aver_dice in this paper improves by 2~3% compared with the control model.

Precise medical image segmentation is important for automating diagnosis and treatment planning in healthcare. While images present the most significant information for segmenting organs using deep learning models, text reports also provide complementary details that can be leveraged to improve segmentation precision. Performance improvement depends on the proper utilization of text reports and the corresponding images. Most attention modules focus on single-modality computation of spatial, channel, or pixel-level attention. They are ineffective in cross-modal alignment, raising issues in multi-modal scenarios. This study addresses these gaps by presenting a text-assisted vision (TAV) model for medical image segmentation with a novel attention computation module named tri-guided attention module (TGAM). TGAM computes visual-visual, language-language, and language-visual attention, enabling the model to understand the important features and correlation between images and medical notes. This module helps the model identify the relevant features within images, text annotations, and text annotations to visual interactions. We incorporate an attention gate (AG) that modulates the influence of TGAM, ensuring it does not overflow the encoded features with irrelevant or redundant information, while maintaining their uniqueness. We evaluated the performance of TAV on two popular datasets containing images and corresponding text annotations. We find TAV to be a new state-of-the-art model, as it improves the performance by 2-7% compared to other models. Extensive experiments were performed to demonstrate the effectiveness of each component of the proposed model.

This study proposes a local bias field and difference estimation (LBDE) model for medical image segmentation and bias field correction. Firstly, the LBDE model uses a linear combination of a given set of smooth orthogonal basis functions, which is called Chebyshev polynomial, to estimate the bias field. Then, a clustering criterion function is defined by considering the difference between the measured image and approximated image in a small region. By applying this difference in the local region, the LBDE model can obtain accurate segmentation results and estimation of the bias field. Finally, the energy functional is incorporated into a level set formulation with a regularisation term, and it is minimised via the level set evolution process. The LBDE model first appears as a two-phase model and then extends to the multi-phase one. Extensive experiments on medical images demonstrate that the LBDE model achieves more precise segmentation results in terms of Jaccard similarity and dice similarity coefficient than the comparative models. Therefore the proposed model can increase the segmentation accuracy and robustness to noise.

In this paper, a novel hybrid active contour model for medical image segmentation is proposed, which integrates the global information of image and Legendre level set. It is a region-based segmentation approach, in which the illumination of the regions of interest is represented by a set of Legendre basis functions in a lower dimensional subspace. Firstly, we present a framework which generalizes the Chan-Vese model and segmentation method based on Legendre level set. The weighting parameter is introduced to control the effect of global and local term on the total energy functional. Secondly, a corresponding termination criterion is employed to ensure the evolving curve automatically stops on true boundaries of objects. Thirdly, experiment results on medical images demonstrate that our method is less sensitive to the initial contour and effective to segment images with inhomogeneous intensity distributions.

Vision foundation models have shown great potential in improving generalizability and data efficiency, especially for medical image segmentation since medical image datasets are relatively small due to high annotation costs and privacy concerns. However, current research on foundation models predominantly relies on transformers. The high quadratic complexity and large parameter counts make these models computationally expensive, limiting their potential for clinical applications. In this work, we introduce Swin-UMamba†, a novel Mamba-based model for medical image segmentation that seamlessly leverages the power of the vision foundation model, which is also computationally efficient with the linear complexity of Mamba. Moreover, we investigated and verified the impact of the vision foundation model on medical image segmentation, in which a self-supervised model adaptation scheme was designed to bridge the gap between natural and medical data. Notably, Swin-UMamba† outperforms 7 state-of-the-art methods, including CNN-based, transformer-based, and Mamba-based approaches across AbdomenMRI, Encoscopy, and Microscopy datasets. The code and models are publicly available at: https://github.com/JiarunLiu/Swin-UMamba.

Medical SAM adapter: Adapting segment anything model for medical image segmentation.

One of the most important areas in medical image analysis is segmentation, in which raw image data is partitioned into structured and meaningful regions to gain further insights. By using Deep Neural Networks (DNN), AI-based automated segmentation algorithms can potentially assist physicians with more effective imaging-based diagnoses. However, since it is difficult to acquire high-quality ground truths for medical images and DNN hyperparameters require significant manual tuning, the results by DNN-based medical models might be limited. A potential solution is to combine multiple DNN models using ensemble learning. We propose a two-layer ensemble of deep learning models in which the prediction of each training image pixel made by each model in the first layer is used as the augmented data of the training image for the second layer of the ensemble. The prediction of the second layer is then combined by using a weight-based scheme which is found by solving linear regression problems. To the best of our knowledge, our paper is the first work which proposes a two-layer ensemble of deep learning models with an augmented data technique in medical image segmentation. Experiments conducted on five different medical image datasets for diverse segmentation tasks show that proposed method achieves better results in terms of several performance metrics compared to some well-known benchmark algorithms. Our proposed two-layer ensemble of deep learning models for segmentation of medical images shows effectiveness compared to several benchmark algorithms. The research can be expanded in several directions like image classification.

A seepage flow model for medical image segmentation has been presented, which linked to the natural phenomenon of "water seeks its own level". The seepage flow to form the segmented pixel set by submerging all the pixels that are r-connected to the initial spring head pixels and also fall within the submerging threshold limits, the moving water fronts keep adding to the segmented pixel set all the pixels until no more pixels fall within the submerging criterion. Based on Mumford-Shah function. A new merit function for criterion has been produced which possess two main kind of measurement reflecting the characteristic of region and contour respectively. Examples are presented to demonstrate the efficiency the model on clinical images segmentation.

Our study briefly discusses the architectures of convolutional neural networks (CNN), their advantages and disadvantages. The features of the architecture of the convolutional neural network U-net are described. An analysis of the CNN U-net was carried out, based on the analysis, a rationale was given for choosing the CNN U-net as the main architecture for using and building subsequent created and analyzed models of cert neural networks to solve the problem of segmentation of medical images. The analysis of architectures of convolutional neural networks, which can be used as convolutional layers in CNN U-net, has been carried out. Based on the analysis, three architectures of convolutional neural networks were selected and described suitable for use as convolutional layers in CNN U-net. Using CNN U-net and three selected convolutional neural networks (“resnet34”, “inceptionv3” and “vgg16”), three neural network models for medical image segmentation were created. The training and testing of the created models of neural networks was carried out. Based on the results of training and testing, an analysis of the obtained indicators was carried out. Experiments were carried out with each of the three constructed models (segmentation of images from the validation set was performed and segmented images were presented). Based on the testing indicators and empirical data obtained from the results of the experiments, the most suitable neural network model created for solving the problem of medical image segmentation was determined. The algorithm for segmentation of medical images has been improved. An algorithm is described that uses the predictions of all created models of neural networks, which demonstrated a more accurate result than each of the considered models separately.

In the realm of medical image segmentation, both CNN-based and Transformer-based models have been extensively explored. However, CNNs exhibit limitations in long-range modeling capabilities, whereas Transformers are hampered by their quadratic computational complexity. Recently, State Space Models (SSMs), exemplified by Mamba, have emerged as a promising approach. They not only excel in modeling long-range interactions but also maintain a linear computational complexity. In this paper, leveraging state space models, we propose a U-shape architecture model for medical image segmentation, named Vision Mamba UNet (VM-UNet). Specifically, the Visual State Space (VSS) block is introduced as the foundation block to capture extensive contextual information, and an asymmetrical encoder-decoder structure is constructed with fewer convolution layers to save calculation cost. We conduct comprehensive experiments on the ISIC17, ISIC18, and Synapse datasets, and the results indicate that VM-UNet performs competitively in medical image segmentation tasks, e.g. obtaining 89.03, 89.71 and 81.08 in terms of DSC score on three datasets respectively. To our best knowledge, this is the first medical image segmentation model constructed based on the pure SSM-based model. We aim to establish a baseline and provide valuable insights for the future development of more efficient and effective SSM-based segmentation systems. Our code is available at https://github.com/JCruan519/VM-UNet .