Domain-Generalized Discrete Diffusion Model for Cross-Domain Medical Image Segmentation.

Medical image segmentation is vital for clinical diagnosis, yet deep learning models face domain shift challenges when test data distributions differ from training data. Single-source domain generalization (DG) has emerged as a solution to these limitations of deep learning models by training models on a single source domain to generalize to unseen target domains. Current single-source DG methods employ domain randomization (e.g., input/feature-space perturbations) to simulate unseen target domains, but face two key limitations: (1) Texture bias in CNN-based architectures due to their local processing nature, leading to overfitting and poor generalization; (2) Restricted augmented style diversity caused by solely source-dependent feature perturbations and input- or feature-only augmentation, resulting in insufficient coverage of target domain distributions and degraded model's generalizationability. Motivated by the Fourier transform's capability to model global dependency and decouple domain invariant features, we propose Random Domain Generalization (RandDG), a frequency-aware domain randomization method for single-source DG in medical image segmentation. This method enhances generalization ability through coordinated input and feature spaces perturbations while maintaining parameter efficiency via lightweight architectural design in the frequencydomain. The proposed RandDG incorporates a novel Global U-Shape Network (GUNet) segmentation architecture for efficient long-range dependency modeling and a Uniform Low Frequency spectrum Transform (ULoFT) filter for feature-space perturbation to enhance domain generalization ability, facilitating the learning of global domain invariant structural information through consistency constraints within the dual-space randomization framework. More specifically, first, the GUNet architecture combines a lightweight encoder with frequency-domain global filtering and forms a U-shape structure through a skip-connected decoder, utilizing Fourier transforms to capture long-range dependencies while minimizing parameters. Second, the ULoFT filter perturbs feature-space styles by Bernoulli mixing statistics of the source low-frequency amplitude spectrum with uniformly sampled values, thereby reducing sole reliance on source statistics and broadening the coverage of potential target domains. Third, a dual-space randomization framework simultaneously applies input-space style augmentation via global intensity nonlinear augmentation (GIN) filters and feature-space amplitude spectrum perturbation via ULoFT filter, regularized by a consistency loss to enforce domain invariant representationlearning. Experiments were conducted on an abdominal CT-MRI dataset for abdominal multi-organ segmentation from CT to T2 spectral presaturation with inversion recovery (T2-SPIR) MRI and a cross-center prostate dataset for prostate segmentation from one center to others. The proposed RandDG method demonstrated superior generalization ability, significantly improving robustness against domain shifts compared to competitive methods. When tested on unseen domains, the proposed method achieved an average DSC of 87.96% and an average HD of 4.82 mm on the abdominal CT-MRI dataset, as well as an average DSC of 75.95% and an average HD of 8.36 mm on the cross-center prostate dataset. Compared to the UNet baseline, the proposed method achieved an improvement of 10.65% in average DSC and a reduction of 5.23 mm in average HD on the abdominal CT-MRI dataset, as well as 19.36% higher average DSC and 3.39 mm lower average HD on the cross-center prostate dataset. Ablation studies confirmed the effectiveness of each component of the proposed RandDGmethod. The proposed RandDG method effectively addresses the critical challenges of texture bias and limited style diversity in single-source DG for medical image segmentation through frequency-aware dual-space randomization framework. This method shows promise for practical deployment in clinical settings, where target domain data may not always beavailable.

Dynamic domain generalization for medical image segmentation.

Generalising deep models to new data from new centres (termed here domains) remains a challenge. This is largely attributed to shifts in data statistics (domain shifts) between source and unseen domains. Recently, gradient-based meta-learning approaches where the training data are split into meta-train and meta-test sets to simulate and handle the domain shifts during training have shown improved generalisation performance. However, the current fully supervised meta-learning approaches are not scalable for medical image segmentation, where large effort is required to create pixel-wise annotations. Meanwhile, in a low data regime, the simulated domain shifts may not approximate the true domain shifts well across source and unseen domains. To address this problem, we propose a novel semi-supervised meta-learning framework with disentanglement. We explicitly model the representations related to domain shifts. Disentangling the representations and combining them to reconstruct the input image allows unlabeled data to be used to better approximate the true domain shifts for meta-learning. Hence, the model can achieve better generalisation performance, especially when there is a limited amount of labeled data. Experiments show that the proposed method is robust on different segmentation tasks and achieves state-of-the-art generalisation performance on two public benchmarks.

Cold SegDiffusion: A novel diffusion model for medical image segmentation

With various face presentation attacks emerging continually, face anti-spoofing (FAS) approaches based on domain generalization (DG) have drawn growing attention. Existing DG-based FAS approaches always capture the domain-invariant features for generalizing on the various unseen domains. However, they neglect individual source domains' discriminative characteristics and diverse domain-specific information of the unseen domains, and the trained model is not sufficient to be adapted to various unseen domains. To address this issue, we propose an Adaptive Mixture of Experts Learning (AMEL) framework, which exploits the domain-specific information to adaptively establish the link among the seen source domains and unseen target domains to further improve the generalization. Concretely, Domain-Specific Experts (DSE) are designed to investigate discriminative and unique domain-specific features as a complement to common domain-invariant features. Moreover, Dynamic Expert Aggregation (DEA) is proposed to adaptively aggregate the complementary information of each source expert based on the domain relevance to the unseen target domain. And combined with meta-learning, these modules work collaboratively to adaptively aggregate meaningful domain-specific information for the various unseen target domains. Extensive experiments and visualizations demonstrate the effectiveness of our method against the state-of-the-art competitors.

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.

The automatic segmentation of medical images has widespread applications in modern clinical workflows. The Segment Anything Model (SAM), a recent development of foundational models in computer vision, has become a universal tool for image segmentation without the need for specific domain training. However, SAM's reliance on prompts necessitates human-computer interaction during the inference process. Its performance on specific domains can also be limited without additional adaptation. In contrast, traditional models like nnUNet are designed to perform segmentation tasks automatically during inference and can work well for each specific domain, but they require extensive training on domain-specificdatasets. To leverage the advantages of both foundational and domain-specific models and achieve fully automated segmentation with limited training samples, we propose nnSAM, which combines the robust feature extraction capabilities of SAM with the automatic configuration abilities of nnUNet to enhance the accuracy and robustness of medical image segmentation on smalldatasets. We propose the nnSAM model for small sample medical image segmentation. We made optimizations for this goal via two main approaches: first, we integrated the feature extraction capabilities of SAM with the automatic configuration advantages of nnUNet, which enables robust feature extraction and domain-specific adaptation on small datasets. Second, during the training process, we designed a boundary shape supervision loss based on level set functions and curvature calculations, enabling the model to learn anatomical shape priors from limited annotationdata. We conducted quantitative and qualitative assessments on the performance of our proposed method on four segmentation tasks: brain white matter, liver, lung, and heart segmentation. Our method achieved the best performance across all tasks. Specifically, in brain white matter segmentation using 20 training samples, nnSAM achieved the highest DICE score of 82.77 ( 10.12) % and the lowest average surface distance (ASD) of 1.14 ( 1.03) mm, compared to nnUNet, which had a DICE score of 79.25 ( 17.24) % and an ASD of 1.36 ( 1.63) mm. A sample size study shows that the advantage of nnSAM becomes more prominent under fewer trainingsamples. A comprehensive evaluation of multiple small-sample segmentation tasks demonstrates significant improvements in segmentation performance by nnSAM, highlighting the vast potential of small-samplelearning.

Deep learning techniques have achieved significant advancements in medical image segmentation in recent years. However, model generalization remains severely constrained by domain shift, particularly in cross-modal medical image segmentation tasks. Traditional segmentation models struggle to generalize effectively to unseen target domains due to differences in joint probability distributions across medical imaging modalities. Existing methods primarily focus on unsupervised domain adaptation (UDA) and domain generalization (DG) techniques. While UDA methods face practical limitations due to difficulties in obtaining target domain data, current DG approaches often overlook inherent anatomical priors in medical images as well as the heterogeneity and sparsity of lesion regions. To address these challenges, this paper proposes a cross-modal medical image segmentation framework that integrates the Vision Mamba model with dynamic domain generalization. The framework achieves cross-domain feature alignment and multi-scale feature fusion by leveraging bidirectional state-space sequence modeling, Bezier curve-style enhancement, and a dual-normalization strategy. Additionally, the VEBlock module is introduced, which effectively combines the dynamic sequence modeling capabilities of the Mamba model with non-local attention mechanisms to better capture cross-modal global dependencies. Experimental results on the BraTS 2018 and cross-modal cardiac datasets demonstrate significant improvements in cross-modal segmentation. For example, in T2 → T1 tasks, our framework achieves an average Dice score of 56.22%, outperforming baseline methods by 1.78% while reducing Hausdorff distance for tumor boundaries to 13.26mm. Furthermore, in the Cardiac CT → MRI tasks, the Hausdorff distance is optimized to 27.34mm, validating the proposed framework's strong generalization capability for complex anatomical structures.

Domain Adaptation (DA) is important for a segmentation model to deal with domain shift in a new target domain. Due to the privacy concern of medical data and the expensive annotation process, Source-Free Domain Adaptation (SFDA) is appealing without access to source data and labels of target domain images for the adaptation. However, existing SFDA methods have limited performance due to insufficient supervision and unreliable pseudo labels. In this paper, we propose an enhanced Iterative Pseudo Label Correction (IPLC+) SFDA framework guided by Segment Anything Model (SAM) for medical image segmentation. Specifically, with a pre-trained source model and SAM, we propose a Reliable SAM Pseudo-label Generator (RSPG) to obtain high-quality and reliable pseudo labels in the target domain based on multiple prompts randomly sampled from the model's prediction. To provide more efficient constraints during adaptation, we introduce self-training pseudo labels weighted by the uncertainty, and propose regularization using mean curvature minimization based on shape-prior knowledge for smoother segmentation. We also propose an Iterative Correction Learning (ICL) strategy to iteratively refine pseudo labels using SAM with updated prompts and combine supervisions to optimize the model sufficiently. Experiments on two public multi-site datasets for prostate and heart segmentation show that our method effectively outperformed ten state-of-the-art SFDA methods, improved the quality of pseudo labels, and even achieved better results than fully supervised learning in the target domain in some cases.

Medical image segmentation tasks hitherto have achieved excellent progresses with large-scale datasets, which empowers us to train potent deep convolutional neural networks (DCNNs). However, labeling such large-scale datasets is laborious and error-prone, which leads the noisy (or incorrect) labels to be an ubiquitous problem in the real-world scenarios. In addition, data collected from different sites usually exhibit significant data distribution shift (or domain shift). As a result, noisy label and domain shift become two common problems in medical imaging application scenarios, especially in medical image segmentation, which degrade the performance of deep learning models significantly. In this paper, we identify a novel problem hidden in medical image segmentation, which is unsupervised domain adaptation on noisy labeled data, and propose a novel algorithm named "Self-Cleansing Unsupervised Domain Adaptation" (S-CDUA) to address such issue. S-CUDA sets up a realistic scenario to solve the above problems simultaneously where training data (i.e., source domain) not only shows domain shift w.r.t. unsupervised test data (i.e., target domain) but also contains noisy labels. The key idea of S-CUDA is to learn noise-excluding and domain invariant knowledge from noisy supervised data, which will be applied on the highly corrupted data for label cleansing and further data-recycling, as well as on the test data with domain shift for supervised propagation. To this end, we propose a novel framework leveraging noisy-label learning and domain adaptation techniques to cleanse the noisy labels and learn from trustable clean samples, thus enabling robust adaptation and prediction on the target domain. Specifically, we train two peer adversarial networks to identify high-confidence clean data and exchange them in companions to eliminate the error accumulation problem and narrow the domain gap simultaneously. In the meantime, the high-confidence noisy data are detected and cleansed in order to reuse the contaminated training data. Therefore, our proposed method can not only cleanse the noisy labels in the training set but also take full advantage of the existing noisy data to update the parameters of the network. For evaluation, we conduct experiments on two popular datasets (REFUGE and Drishti-GS) for optic disc (OD) and optic cup (OC) segmentation, and on another public multi-vendor dataset for spinal cord gray matter (SCGM) segmentation. Experimental results show that our proposed method can cleanse noisy labels efficiently and obtain a model with better generalization performance at the same time, which outperforms previous state-of-the-art methods by large margin. Our code can be found at https://github.com/zzdxjtu/S-cuda.

This study aimed to find suitable source domain data in cross-domain transfer learning to extract robust image features. Then, a model was built to preoperatively distinguish lung granulomatous nodules (LGNs) from lung adenocarcinoma (LAC) in solitary pulmonary solid nodules (SPSNs). Data from 841 patients with SPSNs from five centres were collected retrospectively. First, adaptive cross-domain transfer learning was used to construct transfer learning signatures (TLS) under different source domain data and conduct a comparative analysis. The Wasserstein distance was used to assess the similarity between the source domain and target domain data in cross-domain transfer learning. Second, a cross-domain transfer learning radiomics model (TLRM) combining the best performing TLS, clinical factors and subjective CT findings was constructed. Finally, the performance of the model was validated through multicentre validation cohorts. Relative to other source domain data, TLS based on lung whole slide images as source domain data (TLS-LW) had the best performance in all validation cohorts (AUC range: 0.8228-0.8984). Meanwhile, the Wasserstein distance of TLS-LW was 1.7108, which was minimal. Finally, TLS-LW, age, spiculated sign and lobulated shape were used to build the TLRM. In all validation cohorts, The AUC ranges were 0.9074-0.9442. Compared with other models, decision curve analysis and integrated discrimination improvement showed that TLRM had better performance. The TLRM could assist physicians in preoperatively differentiating LGN from LAC in SPSNs. Furthermore, compared with other images, cross-domain transfer learning can extract robust image features when using lung whole slide images as source domain data and has a better effect.

Domain generalizable (DG) person re-identification (ReID) aims to test across unseen domains without access to the target domain data at training time, which is a realistic but challenging problem. In contrast to methods assuming an identical model for different domains, Mixture of Experts (MoE) exploits multiple domain-specific networks for leveraging complementary information between domains, obtaining impressive results. However, prior MoE-based DG ReID methods suffer from a large model size with the increase of the number of source domains, and most of them overlook the exploitation of domain-invariant characteristics. To handle the two issues above, this paper presents a new approach called Mimic Embedding via adapTive Aggregation (\(\textsf{META}\)) for DG person ReID. To avoid the large model size, experts in \(\textsf{META}\) do not adopt a branch network for each source domain but share all the parameters except for the batch normalization layers. Besides multiple experts, \(\textsf{META}\) leverages Instance Normalization (IN) and introduces it into a global branch to pursue invariant features across domains. Meanwhile, \(\textsf{META}\) considers the relevance of an unseen target sample and source domains via normalization statistics and develops an aggregation module to adaptively integrate multiple experts for mimicking unseen target domain. Benefiting from a proposed consistency loss and an episodic training algorithm, \(\textsf{META}\) is expected to mimic embedding for a truly unseen target domain. Extensive experiments verify that \(\textsf{META}\) surpasses state-of-the-art DG person ReID methods by a large margin. Our code is available at https://github.com/xbq1994/META. KeywordsDomain generalizationPerson re-identification

Eliminating domain-related confounding factors in cross-domain one-shot medical image segmentation via causal inference.

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.

Convolutional neural networks (CNNs), as a typical deep learning technique, have been widely used in image segmentation, but they often require a large amount of annotated data. However, the number of available pixel-wise labeled medical images is extremely small, and this prevents the application of CNNs in many medical image segmentation tasks. We proposed a self-paced learning with diversity SPLD) framework to boost the performances of medical image segmentation models with a limited amount of annotated data. Self-paced learning (SPL) is a learning regime that selects training samples in order from the easiest to the most difficult for model training. In addition, we took the diversity of the data into consideration. The proposed self-paced learning with diversity by query-by-committee (SPLD-QBC) algorithm dynamically and diversely selects the appropriate training data to boost the performance of an image segmentation model. SPLD-QBC incorporates the query-by-committee (QBC) technique for data selection and affinity propagation for optimizing the data diversity. By dynamically selecting the optimal sequence of training samples from different probability distributions, the segmentation models achieved improved performances. To verify the effectiveness of the proposed SPLD-QBC framework, we conducted experiments on three medical image segmentation tasks with five different datasets. The experimental results indicated that the proposed SPLD-QBC algorithm significantly improved upon the segmentation performances of the baseline models and resulted in a higher Dice score, surface distance and mean intersection over union (mIoU). The proposed SPLD significantly boosts the segmentation performances of models and is easily embedded into CNN-based image segmentation models.