Multiscale fused network with additive channel–spatial attention for image segmentation

Medical image segmentation is a critical task in computer-assisted diagnosis and therapy planning. Accurate and efficient fully automated segmentation methods are essential for accelerating decision-making processes. Compared with natural image segmentation methods, current deep learning–based segmentation models for medical imaging face challenges related to data scarcity, inconsistent annotations, and high computational cost. Consequently, this thesis aims to develop efficiency and generalised medical image segmentation models, which are fundamental to the clinical adoption and industrial deployment of this technology in healthcare settings. This thesis systematically addresses these challenges through a series of progressively advanced deep learning models. The foundational work, CTranS, forms a baseline medical image segmentation model that combines a convolutional neural network (CNN) based encoder with a Transformer-based decoder. This initial contribution leverages the strengths of CNNs for local feature extraction and Transformers for global contextual understanding, achieving state-of-the-art (SOTA) performance. Subsequently, to enhance the generalisability, MO-CTranS is introduced, which is capable of learning from a mixture of multiple partially labelled datasets. MO-CTranS employs a CNN encoder and Transformer decoder with task-specific tokens to handle label inconsistencies and class imbalances, tackling the challenges of training a single unified model from fragmented and heterogeneously annotated medical data. This innovation moves beyond single-task, fully-annotated model training towards a more scalable and clinically realistic approach. Despite the superior performance of CTranS and MO-CTranS, the computational efficiency of the Transformer blocks hinders their clinical application. Thus, CRFTrans is proposed to enhance the model efficiency. As a novel recursive Vision Transformer layer designed to replace the commonly used cascaded Transformer layers, this approach represents a fundamental shift from simply stacking layers to a recursive refinement process, demonstrating that comparable performance can be achieved with significantly reduced computational complexity, thus facilitating deployment in resource-constrained clinical environments. However, these methods still depend on extensive manual annotation, which remains challenging and time-consuming in clinical practice. Thus, to further enhance both efficiency and generality, a novel semi-supervised medical image segmentation framework, referred to as SSL-MedSAM2, has been developed. SSL-MedSAM2 integrates a training-free few-shot segmentation branch for pseudo labelling and a fully supervised learning branch for label refinement. This design allows the model to leverage a small fraction of labelled images together with a large unlabelled set, dramatically decreasing the annotation effort. Together, these contributions result in deep learning models that are significantly more data-efficient and broadly applicable than conventional approaches. The overall impact of this work is to facilitate the reliable and scalable deployment of segmentation models in real-world clinical settings by bridging the gap among accuracy, efficiency, and generalisability.

Introduction: Precise segmentation of the liver from computed tomography (CT) scans is an essential component in the diagnosis and treatment of liver diseases such as hepatocellular carcinoma. Manual segmentation can be time-consuming and subject to variability, spurring the application of automated systems based on deep learning. We introduce a 3D U-Net-based convolutional neural network for end-to-end segmentation of the liver from volumetric CT images. The model is trained and tested on the LiTS dataset using a standard preprocessing pipeline and an in-house composite loss function—AdvancedComboLoss—combining Dice Loss, Binary Cross-Entropy, and Focal Tversky Loss to balance class imbalance and boundary accuracy. The network handles entire 3D volumes with spatial context retained between anatomical planes and is trained on a GPU-enabled setup with the PyTorch environment. Objectives: To develop a highly accurate and computationally efficient 3D U-Net-based deep learning model for automated liver segmentation from volumetric CT scans, addressing challenges like class imbalance, anatomical variability, and low tissue contrast, using a custom loss function (AdvancedComboLoss) and a standardized preprocessing pipeline. Methods: The methodology adopted is volumetric liver segmentation from a specially designed 3D U-Net architecture specific to medical CT scans. The process entails preprocessing of raw CT scans into a common shape and intensity range and training a fully convolutional 3D neural network that learns the spatial relationships among all three anatomical planes. The model uses an encoder-decoder architecture with skip connections to maintain fine-grained structural information, allowing for accurate delineation of liver regions. A hybrid loss function—integrating Dice, Binary Cross-Entropy, and Focal Tversky losses—is used to direct the learning process to optimize overlap of regions, accuracy of classification, and class imbalance. Results: Experimental outcomes demonstrate that the model has a Dice score of 0.9604 and voxel-wise accuracy of 0.9979, indicating excellent performance in segmentation as well as strong agreement with expert annotations. The model presented in this paper presents a solid, scalable, and clinically valuable approach to automated liver segmentation with future potential to integrate into multi-stage liver and tumor analysis workflows. Conclusions: Herein, we built and tested a 3D U-Net deep learning architecture for unsupervised liver segmentation from abdominal CT volumes. Through utilization of volumetric convolutions as well as through a well-engineered composite loss function, the model was successful in capturing complicated liver anatomy as well as being able to hold high performance even on varied CT volumes. The power of the method is not just in its quantitative performance but also in its capability to generate anatomically consistent and visually plausible segmentations, making it extremely relevant to actual clinical workflows. Although there are some limitations—constrained batch size and absence of data augmentation—there is room for improvement, the findings affirm that deep learning can be an effective tool for automating key tasks in medical image analysis. As a whole, this paper adds a robust and effective segmentation model that would be potentially embedded in wider diagnostic workflows, ending up improving the accuracy, speed, and scalability of liver evaluation in clinical practice. In brief, this paper adds a high-fidelity, scalable, and computationally effective solution for computer-aided liver segmentation, pushing the application of artificial intelligence in clinical decision-making and medical image analysis.

Segmentation of intracranial aneurysm images acquired using magnetic resonance angiography (MRA) is essential for medical auxiliary treatments, which can effectively prevent subarachnoid hemorrhages. This paper proposes an image segmentation model based on a dense convolutional attention U-Net, which fuses deep and rich semantic information with shallow-detail information for adaptive and accurate segmentation of MRA-acquired aneurysm images with large size differences. The U-Net model serves as a backbone, combining dense block and convolution block attention module (CBAM). The dense block is composed of a batch normalization layer, an randomly rectified linear unit activation function, and a convolutional layer, for mitigation of vanishing gradients, for multiplexing of aneurysm features, and for improving the network training efficiency. The CBAM is composed of a channel attention module and a spatial attention module, improving the segmentation performance of feature discrimination and enhancing the acquisition of key feature information. Owing to the large variation of aneurysm sizes, multi-scale fusion is performed during up-sampling, for adaptive segmentation of MRA-acquired aneurysm images. The model was tested on the MICCAI 2020 ADAM dataset, and its generalizability was validated on the clinical aneurysm dataset (aneurysm sizes: < 3 mm, 3–7 mm, and > 7 mm) supplied by the Affiliated Hospital of Qingdao University. A good clinical application segmentation performance was demonstrated.

Medical image segmentation is a critical and important step for developing computer-aided system in clinical situations. It remains a complicated and challenging task due to the large variety of imaging modalities and different cases. Recently, Unet has become one of the most popular deep learning frameworks because of its accurate performance in biomedical image segmentation. In this paper, we propose a contour-aware semantic segmentation network, which is an extension of Unet, for medical image segmentation. The proposed method includes a semantic branch and a detail branch. The semantic branch focuses on extracting the semantic features from shallow and deep layers; the detail branch is used to enhance the contour information implied in the shallow layers. In order to improve the representation capability of the network, a MulBlock module is designed to extract semantic information with different receptive fields. Spatial attention module (CAM) is used to adaptively suppress the redundant features. In comparison with the state-of-the-art methods, our method achieves a remarkable performance on several public medical image segmentation challenges.

Medical-image segmentation is of primary importance in the development of computer assisted detection (CAD) of mammographic masses. Image segmentation is used for finding regions of interest (ROI) within mammograms that are further processed to estimate their likelihood of being a potential lesion or mass. This task can be stated as follows: €œA gray-scale image is represented as a 1D array X = (x1, x2,...,xN), where xn is an input feature for pixel n and N is the total number of pixels in the image. The input feature vector xn may be a D dimensional vector or simply the gray-scale value of the pixel n. Let the underlying true segmentation of the image be denoted as Y = (y1, y2,..., yN). It is assumed that the number of classes is predetermined as a set of known class labels I‰l, where l Iµ {1,..., L{ and therefore the class label of pixel n is indicated as y n ∈{I‰ l } L l=1 . The task is to predict the A·n estimate of the segmentation for each pixel. In an automated system, such likelihood estimates are generated using a classifier that takes as input some features that describe region properties, e.g., image texture. The segmentation of digital mammograms is a difficult task. A number of different approaches have been adopted in this area with varying success. Broadly speaking, image segmentation can be carried out either as: (1) image preprocessing step with the main objective of identifying regions of interest (spatially coherent collection of pixels within the image) and a subsequent feature extraction and classification step is required to characterize a given region of interest as either €œnormal€ or €œsuspicious€; or (2) part of the overall classification scheme, where the likelihood of each pixel being €œsuspicious is calculated to generate a probability image using pixel-based features as input to a classifier. The probability image is thereafter thresholded to give both the segmentation and classification output. The main difference between (1) and (2) is that in the first case, image segmentation is needed before classification can be performed; whereas in the second case, classification and segmentation are both based on thresholding probability estimates of pixels and are achieved at the same time. The image segmentation approaches, with the aim of finding regions of interest, were first attempted in digital mammography with varying techniques, including global thresholding, region growing, region clustering, template matching, edge operators and filtering, and edge-based hybrid methods. Image segmentation is developed from pixel-based probability estimation is underpinned by modeling features using parametric methods, Gaussian mixture modeling, Markov random fields, and other methods. A detailed discussion on mammogram-based image segmentation with their algorithms is available in Singh and Bovis. Experience shows that it is quite difficult to select an outright winning image-segmentation method, since different approaches have been shown to be suitable for different data sets and show varied performance under different experimental conditions. However, there is a consensus that image segmentation in medical imaging is most successful when explicitly using a priori information about the problem to be studied.

– Medical image segmentation is considered the most precarious element in the analysis and processing of real-life images in the clinical sector. Actually, segmentation effects affect the subsequent procedures of image evaluation, life object illustration and description, feature dimensions and the subsequent considerable task levels such as object categorization. In that case, medical image segmentation is the most essential and crucial aspect for aiding the visualization, delineation and depiction of regions of interest for any particular image. The aspect of physical segmentation of a picture is not just challenging and time-consuming task to do, but also problematic to assure accuracy considering the wide-range image modalities and unguided image quantities which have to be observed. In that case, it now becomes fundamental to evaluate the present approaches of image segmentation based on the application of computing algorithms which require the interaction of users with evaluating medical images. In the process of image segmentation, an anatomic classification requires to be extracted or defined to be projected effectively and independently. In that case, this contribution is focused on image segmentation to extract details for decision-making in the clinical sector. The paper presents generalized and relative techniques that have been categorized into three groups: pixel-centre, edge-centred and region-centred techniques. The paper also provides a highlight of the strengths and weaknesses of these techniques in reference to the appropriateness of a wide-range application in medical image segmentation.

Deep learning techniques have rapidly become important as a preferred method for evaluating medical image segmentation. This survey analyses different contributions in the deep learning medical field, including the major common issues published in recent years, and also discusses the fundamentals of deep learning concepts applicable to medical image segmentation. The study of deep learning can be applied to image categorization, object recognition, segmentation, registration, and other tasks. First, the basic ideas of deep learning techniques, applications, and frameworks are introduced. Deep learning techniques that operate the ideal applications are briefly explained. This paper indicates that there is a previous experience with different techniques in the class of medical image segmentation. Deep learning has been designed to describe and respond to various challenges in the field of medical image analysis such as low accuracy of image classification, low segmentation resolution, and poor image enhancement. Aiming to solve these present issues and improve the evolution of medical image segmentation challenges, we provide suggestions for future research.

7 - A survey on neutrosophic medical image segmentation

Image segmentation is the process of delineating the regions occupied by objects of interest in a given image. This operation is a fundamentally required first step in numerous applications of medical imagery. In the medical imaging field, this activity has a rich literature that spans over 45 years. In spite of numerous advances, including deep learning (DL) networks (DLNs) in recent years, the problem has defied a robust, fail-safe, and satisfactory solution, especially for objects that are manifest with low contrast, are spatially sparse, have variable shape among individuals, or are affected by imaging artifacts, pathology, or post-treatment change in the body. Although image processing techniques, notably DLNs, are uncanny in their ability to harness low-level intensity pattern information on objects, they fall short in the high-level task of identifying and localizing an entire object as a gestalt. This dilemma has been a fundamental unmet challenge in medical image segmentation. In this paper, we demonstrate that by synergistically marrying the unmatched strengths of high-level human knowledge (i.e., natural intelligence (NI)) with the capabilities of DL networks (i.e., artificial intelligence (AI)) in garnering intricate details, these challenges can be significantly overcome. Focusing on the object recognition task, we formulate an anatomy-guided DL object recognition approach named Automatic Anatomy Recognition-Deep Learning (AAR-DL) which combines an advanced anatomy-modeling strategy, model-based non-DL object recognition, and DL object detection networks to achieve expert human-like performance.

Deep learning has achieved great success in the field of computer vision, and the precision in image classification and image detection has surpassed humans. Therefore, this paper combines deep learning and medical image segmentation, focusing on how to improve the accuracy and speed of segmentation algorithm of medical exercise rehabilitation image. Aiming at the shortcomings of traditional medical image recognition methods, a medical exercise rehabilitation image segmentation algorithm based on hierarchical features of convolutional neural networks is proposed, this paper calls it as hierarchical features of convolutional neural networks (HFCNN). The algorithm firstly samples the convolution output of multiple layers in the convolutional neural network to a unified scale and combines them to construct a hierarchical feature. This hierarchical feature combines the structural information of objects contained in the shallow layer of the network with the semantic information of objects contained in the deep layers of the network, so it has a strong ability to express. Secondly, the image can be segmented into multiple super pixels by the super pixel segmentation algorithm. The classifier is trained using the hierarchical features of the super pixel, and then the classification result is mapped back to the pixel. Finally, a fully connected conditional random field algorithm including one-potential potential energy and paired potential energy is constructed. The corresponding energy function is used to smooth the classification result of the pixel, and the regional consistency and continuity of the pixel mark are improved. Compared with many classical convolutional neural network algorithms, this algorithm not only accelerates the network convergence speed, shortens the training time, but also significantly improves the accuracy of segmentation algorithm of medical exercise rehabilitation image, showing good practical value.

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.

DSEU-net: A novel deep supervision SEU-net for medical ultrasound image segmentation

Ultrasound imaging can distinctly display the morphology and structure of internal organs within the human body, enabling the examination of organs like the breast, liver, and thyroid. It can identify the locations of tumors, nodules, and other lesions, thereby serving as an efficacious tool for treatment detection and rehabilitation evaluation. Typically, the attending physician is required to manually demarcate the boundaries of lesion locations, such as tumors, in ultrasound images. Nevertheless, several issues exist. The high noise level in ultrasound images, the degradation of image quality due to the impact of surrounding tissues, and the influence of the operator’s experience and proficiency on the determination of lesion locations can all contribute to a reduction in the accuracy of delineating the boundaries of lesion sites. In the wake of the advancement of deep learning, its application in medical image segmentation is becoming increasingly prevalent. For instance, while the U-Net model has demonstrated a favorable performance in medical image segmentation, the convolution layers of the traditional U-Net model are relatively simplistic, leading to suboptimal extraction of global information. Moreover, due to the significant noise present in ultrasound images, the model is prone to interference. In this research, we propose an Attention Residual Network model (ARU-Net). By incorporating residual connections within the encoder section, the learning capacity of the model is enhanced. Additionally, a spatial hybrid convolution module is integrated to augment the model’s ability to extract global information and deepen the vertical architecture of the network. During the feature fusion stage of the skip connections, a channel attention mechanism and a multi-convolutional self-attention mechanism are respectively introduced to suppress noisy points within the fused feature maps, enabling the model to acquire more information regarding the target region. Finally, the predictive efficacy of the model was evaluated using publicly accessible breast ultrasound and thyroid ultrasound data. The ARU-Net achieved mean Intersection over Union (mIoU) values of 82.59% and 84.88%, accuracy values of 97.53% and 96.09%, and F1-score values of 90.06% and 89.7% for breast and thyroid ultrasound, respectively.

YoTransViT: A transformer and CNN method for predicting and classifying skin diseases using segmentation techniques

Statistical as well as adaptive clustering approaches are being currently used for both segmentation and vector quantization of medical images. However, a comparative evaluation of both approaches has rarely been done to identify the efficacy of such approaches to specific applications, for example, image segmentation and vector quantization. The rate distortion functions of three clustering algorithms, namely, the statistical based deterministic annealing, the adaptive fuzzy leader clustering algorithm, and LBG, have been computed for vector quantization using multi-scale vectors in the wavelet domain. Such comparative evaluation serves as a guide for proper selection of clustering algorithms for global codebook generation in vector quantization and for image segmentation.