Both CNN-based and Transformer-based methods have achieved remarkable success in medical image segmentation tasks. However, CNN-based methods struggle to effectively capture global contextual information due to the inherent limitations of convolution operations. Meanwhile, Transformer-based methods suffer from insufficient local feature modeling and face challenges related to the high computational complexity caused by the self-attention mechanism. To address these limitations, we propose a novel hybrid CNN-Transformer architecture, named MSLAU-Net, which integrates the strengths of both paradigms. The proposed MSLAU-Net incorporates two key ideas. First, it introduces Multi-Scale Linear Attention, designed to efficiently extract multi-scale features from medical images while modeling long-range dependencies with low computational complexity. Second, it adopts a top-down feature aggregation mechanism, which performs multi-level feature aggregation and restores spatial resolution using a lightweight structure. Extensive experiments conducted on benchmark datasets covering three imaging modalities demonstrate that the proposed MSLAU-Net outperforms other state-of-the-art methods on nearly all evaluation metrics, validating the superiority, effectiveness, and robustness of our approach. Our code is available at https://github.com/Monsoon49/MSLAU-Net.

3D medical image segmentation is vital for clinical diagnosis and treatment but is challenged by high-dimensional data and complex spatial dependencies. Traditional single-modality networks, such as CNNs and Transformers, are often limited by computational inefficiency and constrained contextual modeling in 3D settings. We introduce a novel multimodal framework that leverages Mamba and Kolmogorov-Arnold Networks (KAN) as an efficient backbone for long-sequence modeling. Our approach features three key innovations: First, an EGSC (Enhanced Gated Spatial Convolution) module captures spatial information when unfolding 3D images into 1D sequences. Second, we extend Group-Rational KAN (GR-KAN), a Kolmogorov-Arnold Networks variant with rational basis functions, into 3D-Group-Rational KAN (3D-GR-KAN) for 3D medical imaging - its first application in this domain - enabling superior feature representation tailored to volumetric data. Third, a dual-branch text-driven strategy leverages CLIP's text embeddings: one branch swaps one-hot labels for semantic vectors to preserve inter-organ semantic relationships, while the other aligns images with detailed organ descriptions to enhance semantic alignment. Experiments on the Medical Segmentation Decathlon (MSD) and KiTS23 datasets show our method achieving state-of-the-art performance, surpassing existing approaches in accuracy and efficiency. This work highlights the power of combining advanced sequence modeling, extended network architectures, and vision-language synergy to push forward 3D medical image segmentation, delivering a scalable solution for clinical use. The source code is openly available at https://github.com/yhy-whu/TK-Mamba.

Segmentation of 3D medical images, and brain segmentation in particular, is an important topic in neuroimaging and in radiotherapy. Overcoming the current, time consuming, practise of manual delineation of brain tumours and providing an accurate, explainable, and replicable method of segmentation of the tumour area and related tissues is therefore an open research challenge. In this paper, we first propose a novel symbolic approach to brain segmentation and delineation of brain lesions based on spatial model checking. This method has its foundations in the theory of closure spaces, a generalisation of topological spaces, and spatial logics. At its core is a high-level declarative logic language for image analysis, ImgQL, and an efficient spatial model checker, VoxLogicA, exploiting state-of-the-art image analysis libraries in its model checking algorithm. We then illustrate how this technique can be combined with Machine Learning techniques leading to a hybrid AI approach that provides accurate and explainable segmentation results. We show the results of the application of the symbolic approach on several public datasets with 3D magnetic resonance (MR) images. Three datasets are provided by the 2017, 2019 and 2020 international MICCAI BraTS Challenges with 210, 259 and 293 MR images, respectively, and the fourth is the BrainWeb dataset with 20 (synthetic) 3D patient images of the normal brain. We then apply the hybrid AI method to the BraTS 2020 training set. Our segmentation results are shown to be in line with the state-of-the-art with respect to other recent approaches, both from the accuracy point of view as well as from the view of computational efficiency, but with the advantage of them being explainable.

Stroke is among the top three causes of death worldwide, and accurate identification of stroke lesion boundaries is critical for diagnosis and treatment. Supervised deep learning methods have emerged as the leading solution for stroke lesion segmentation but require large, diverse, and annotated datasets. The ISLES'24 challenge addresses this need by providing longitudinal stroke imaging data, including CT scans taken on arrival to the hospital and follow-up MRI taken 2-9 days from initial arrival, with annotations derived from follow-up MRI. Importantly, models submitted to the ISLES'24 challenge are evaluated using only CT inputs, requiring prediction of lesion progression that may not be visible in CT scans for segmentation. Our winning solution shows that a carefully designed preprocessing pipeline including deep-learning-based skull stripping and custom intensity windowing is beneficial for accurate segmentation. Combined with a standard large residual nnU-Net architecture for segmentation, this approach achieves a mean test Dice of 28.5 with a standard deviation of 21.27.

Medical image segmentation is vital for clinical diagnosis, yet current deep learning methods often demand extensive expert effort, i.e., either through annotating large training datasets or providing prompts at inference time for each new case. This paper introduces a zero-shot and automatic segmentation pipeline that combines off-the-shelf vision-language and segmentation foundation models. Given a medical image and a task definition (e.g., "segment the optic disc in an eye fundus image"), our method uses a grounding model to generate an initial bounding box, followed by a visual prompt boosting module that enhance the prompts, which are then processed by a promptable segmentation model to produce the final mask. To address the challenges of domain gap and result verification, we introduce a test-time adaptation framework featuring a set of learnable adaptors that align the medical inputs with foundation model representations. Its hyperparameters are optimized via Bayesian Optimization, guided by a proxy validation model without requiring ground-truth labels. Our pipeline offers an annotation-efficient and scalable solution for zero-shot medical image segmentation across diverse tasks. Our pipeline is evaluated on seven diverse medical imaging datasets and shows promising results. By proper decomposition and test-time adaptation, our fully automatic pipeline performs competitively with weakly-prompted interactive foundation models.

Accurate lung volume determination is crucial for reliable dark-field imaging. We compared different approaches for the determination of lung volume in mean dark-field coefficient calculation. In this retrospective analysis of data prospectively acquired between October 2018 and October 2020, patients at least 18 years of age who underwent chest computed tomography (CT) were screened for study participation. Inclusion criteria were the ability to consent and to stand upright without help. Exclusion criteria were pregnancy, lung cancer, pleural effusion, atelectasis, air space disease, ground-glass opacities, and pneumothorax. Lung volume was calculated using four methods: conventional radiography (CR) using shape information; a convolutional neural network (CNN) trained for CR; CT-based volume estimation; and results from pulmonary function testing (PFT). Results were compared using a Student t-test and Spearman ρ correlation statistics. We studied 81 participants (51 men, 30 women), aged 64 ± 12 years (mean ± standard deviation). All lung volumes derived from the various methods were different from each other: CR, 7.27 ± 1.64 L; CNN, 4.91 ± 1.05 L; CT, 5.25 ± 1.36 L; PFT, 6.54 L ± 1.52 L; p < 0.001 for all comparisons. A high positive correlation was found for all combinations (p < 0.001 for all), the highest one being between CT and CR (ρ = 0.88) and the lowest one between PFT and CNN (ρ = 0.78). Lung volume and therefore mean dark-field coefficient calculation is highly dependent on the method used, taking into consideration different positioning and inhalation depths. This study underscores the impact of the method used for lung volume determination. In the context of mean dark-field coefficient calculation, CR-based methods are more desirable because both dark-field images and conventional images are acquired at the same breathing state, and therefore, biases due to differences in inhalation depth are eliminated. Lung volume measurements vary significantly between different determination methods. Mean dark-field coefficient calculations require the same method to ensure comparability. Radiography-based methods simplify workflows and minimize biases, making them most suitable.

Low-quality pseudo labels pose a significant obstacle in semi-supervised medical image segmentation (SSMIS), impeding consistency learning on unlabeled data. Leveraging vision-language model (VLM) holds promise in ameliorating pseudo label quality by employing textual prompts to delineate segmentation regions, but it faces the challenge of cross-modal alignment uncertainty due to multiple correspondences (multiple images/texts tend to correspond to one text/image). Existing VLMs address this challenge by modeling semantics as distributions but such distributions lead to semantic degradation. To address these problems, we propose Alignment-Multiplicity Aware Vision-Language Model (AMVLM), a new VLM pre-training paradigm with two novel similarity metric strategies. (i) Cross-modal Similarity Supervision (CSS) proposes a probability distribution transformer to supervise similarity scores across fine-granularity semantics through measuring cross-modal distribution disparities, thus learning cross-modal multiple alignments. (ii) Intra-modal Contrastive Learning (ICL) takes into account the similarity metric of coarse-fine granularity information within each modality to encourage cross-modal semantic consistency. Furthermore, using the pretrained AMVLM, we propose a pioneering text-guided SSMIS network to compensate for the quality deficiencies of pseudo-labels. This network incorporates a text mask generator to produce multimodal supervision information, enhancing pseudo label quality and the model's consistency learning. Extensive experimentation validates the efficacy of our AMVLM-driven SSMIS, showcasing superior performance across four publicly available datasets. The code will be available at: https://github.com/QingtaoPan/AMVLM.

BackgroundThis study aims to present the Segmentation-based Myocardial Advanced Refinement Tracking (SMART) system, a novel artificial intelligence (AI)-based framework for transthoracic echocardiography (TTE) that incorporates motion tracking and left ventricular (LV) myocardial segmentation for automated LV mass (LVM) and global longitudinal strain (LVGLS) assessment. MethodsThe SMART system demonstrates LV speckle tracking based on motion vector estimation, refined by structural information using endocardial and epicardial segmentation throughout the cardiac cycle. This approach enables automated measurement of LVMSMART and LVGLSSMART. The feasibility of SMART is validated in 111 hypertrophic cardiomyopathy (HCM) patients (median age: 58 years, 69% male) who underwent TTE and cardiac magnetic resonance imaging (CMR). ResultsLVGLSSMART showed a strong correlation with conventional manual LVGLS measurements (Pearsons correlation coefficient [PCC] 0.851; mean difference 0 [-2-0]). When compared to CMR as the reference standard for LVM, the conventional dimension-based TTE method overestimated LVM (PCC 0.652; mean difference: 106 [90-123]), whereas LVMSMART demonstrated excellent agreement with CMR (PCC 0.843; mean difference: 1 [-11-13]). For predicting extensive myocardial fibrosis, LVGLSSMART and LVMSMART exhibited performance comparable to conventional LVGLS and CMR (AUC: 0.72 and 0.66, respectively). Patients identified as high-risk for extensive fibrosis by LVGLSSMART and LVMSMART had significantly higher rates of adverse outcomes, including heart failure hospitalization, new-onset atrial fibrillation, and defibrillator implantation. ConclusionsThe SMART technique provides a comparable LVGLS evaluation and a more accurate LVM assessment than conventional TTE, with predictive values for myocardial fibrosis and adverse outcomes. These findings support its utility in HCM management.

The explosive growth of available high-quality imaging data coupled with new progress in hardware capabilities has enabled a new era of unprecedented performance in brain segmentation tasks. Despite the explosion of new data released by consortiums and groups around the world, most published, closed, or openly available segmentation models have either a limited or an unknown role in pediatric brains. This study explores the utility of state-of-the-art automated ventricular segmentation tools applied to pediatric hydrocephalus. Two popular, fast, whole-brain segmentation tools were used (FastSurfer and QuickNAT) to automatically segment the lateral ventricles and evaluate their accuracy in children with hydrocephalus. Forty scans from 32 patients were included in this study. The patients underwent imaging at the University of Minnesota Medical Center or satellite clinics, were between 0 and 18 years old, had an ICD-10 diagnosis that included the word hydrocephalus, and had at least one T1-weighted pre- or postcontrast MPRAGE sequence. Patients with poor quality scans were excluded. Dice similarity coefficient (DSC) scores were used to compare segmentation outputs against manually segmented lateral ventricles. Overall, both models performed poorly with DSCs of 0.61 for each segmentation tool. No statistically significant difference was noted between model performance (p = 0.86). Using a multivariate linear regression to examine factors associated with higher DSC performance, male gender (p = 0.66), presence of ventricular catheter (p = 0.72), and MRI magnet strength (p = 0.23) were not statistically significant factors. However, younger age (p = 0.03) and larger ventricular volumes (p = 0.01) were significantly associated with lower DSC values. A large-scale visualization of 196 scans in both models showed characteristic patterns of segmentation failure in larger ventricles. Significant gaps exist in current cutting-edge segmentation models when applied to pediatric hydrocephalus. Researchers will need to address these types of gaps in performance through thoughtful consideration of their training data before reaching the ultimate goal of clinical deployment.

For patients with colorectal liver metastases (CRLM), total tumor volume (TTV) is prognostic. A deep-learning segmentation model for CRLM to assess TTV called COlorectal cAncer Liver metastases Assessment (COALA) has been developed. This study evaluated COALA's performance and practical utility in the radiological picture archiving and communication system (PACS). A secondary aim was to provide lessons for future researchers on the implementation of artificial intelligence (AI) models. Patients discussed between January and December 2023 in a multidisciplinary meeting for CRLM were included. In those patients, CRLM was automatically segmented in portal-venous phase CT scans by COALA and integrated with PACS. Eight expert abdominal radiologists completed a questionnaire addressing segmentation accuracy and PACS integration. They were also asked to write down general remarks. In total, 57 patients were evaluated. Of those patients, 112 contrast-enhanced portal-venous phase CT scans were analyzed. Of eight radiologists, six (75%) evaluated the model as user-friendly in their radiological workflow. Areas of improvement of the COALA model were the segmentation of small lesions, heterogeneous lesions, and lesions at the border of the liver with involvement of the diaphragm or heart. Key lessons for implementation were a multidisciplinary approach, a robust method prior to model development and organizing evaluation sessions with end-users early in the development phase. This study demonstrates that the deep-learning segmentation model for patients with CRLM (COALA) is user-friendly in the radiologist's PACS. Future researchers striving for implementation should have a multidisciplinary approach, propose a robust methodology and involve end-users prior to model development. Many segmentation models are being developed, but none of those models are evaluated in the (radiological) workflow or clinically implemented. Our model is implemented in the radiological work system, providing valuable lessons for researchers to achieve clinical implementation. Developed segmentation models should be implemented in the radiological workflow. Our implemented segmentation model provides valuable lessons for future researchers. If implemented in clinical practice, our model could allow for objective radiological evaluation.