Scribble-based weakly supervised segmentation methods have shown promising results in medical image segmentation, significantly reducing annotation costs. However, existing approaches often rely on auxiliary tasks to enforce semantic consistency and use hard pseudo labels for supervision, overlooking the unique challenges faced by models trained with sparse annotations. These models must predict pixel-wise segmentation maps from limited data, making it crucial to handle varying levels of annotation richness effectively. In this paper, we propose MaCo, a weakly supervised model designed for medical image segmentation, based on the principle of "from few to more." MaCo leverages Masked Context Modeling (MCM) and Continuous Pseudo Labels (CPL). MCM employs an attention-based masking strategy to perturb the input image, ensuring that the model's predictions align with those of the original image. CPL converts scribble annotations into continuous pixel-wise labels by applying an exponential decay function to distance maps, producing confidence maps that represent the likelihood of each pixel belonging to a specific category, rather than relying on hard pseudo labels. We evaluate MaCo on three public datasets, comparing it with other weakly supervised methods. Our results show that MaCo outperforms competing methods across all datasets, establishing a new record in weakly supervised medical image segmentation.

Although recent years have witnessed significant advancements in medical image segmentation, the pervasive issue of domain shift among medical images from diverse centres hinders the effective deployment of pre-trained models. Many Test-time Adaptation (TTA) methods have been proposed to address this issue by fine-tuning pre-trained models with test data during inference. These methods, however, often suffer from less-satisfactory optimization due to suboptimal optimization direction (dictated by the gradient) and fixed step-size (predicated on the learning rate). In this paper, we propose the Gradient alignment-based Test-time adaptation (GraTa) method to improve both the gradient direction and learning rate in the optimization procedure. Unlike conventional TTA methods, which primarily optimize the pseudo gradient derived from a self-supervised objective, our method incorporates an auxiliary gradient with the pseudo one to facilitate gradient alignment. Such gradient alignment enables the model to excavate the similarities between different gradients and correct the gradient direction to approximate the empirical gradient related to the current segmentation task. Additionally, we design a dynamic learning rate based on the cosine similarity between the pseudo and auxiliary gradients, thereby empowering the adaptive fine-tuning of pre-trained models on diverse test data. Extensive experiments establish the effectiveness of the proposed gradient alignment and dynamic learning rate and substantiate the superiority of our GraTa method over other state-of-the-art TTA methods on a benchmark medical image segmentation task.

Recent advancements in multimodal large language models (MLLMs) have significantly improved performance in visual question answering. However, they often suffer from hallucinations. In this work, hallucinations are categorized into two main types: initial hallucinations and snowball hallucinations. We argue that adequate contextual information can be extracted directly from the token interaction process. Inspired by causal inference in the decoding strategy, we propose to leverage causal masks to establish information propagation between multimodal tokens. The hypothesis is that insufficient interaction between those tokens may lead the model to rely on outlier tokens, overlooking dense and rich contextual cues. Therefore, we propose to intervene in the propagation process by tackling outlier tokens to enhance in-context inference. With this goal, we present FarSight, a versatile plug-and-play decoding strategy to reduce attention interference from outlier tokens merely by optimizing the causal mask. The heart of our method is effective token propagation. We design an attention register structure within the upper triangular matrix of the causal mask, dynamically allocating attention to capture attention diverted to outlier tokens. Moreover, a positional awareness encoding method with a diminishing masking rate is proposed, allowing the model to attend to further preceding tokens, especially for video sequence tasks. With extensive experiments, FarSight demonstrates significant hallucination-mitigating performance across different MLLMs on both image and video benchmarks, proving its effectiveness.

Self-supervised learning (SSL) has long had great success in advancing the field of annotation-efficient learning. However, when applied to CT volume segmentation, most SSL methods suffer from two limitations, including rarely using the information acquired by different imaging modalities and providing supervision only to the bottleneck encoder layer. To address both limitations, we design a pretext task to align the information in each 3D CT volume and the corresponding 2D generated X-ray image and extend self-distillation to deep self-distillation. Thus, we propose a self-supervised learner based on Cross-modal Alignment and Deep Self-distillation (CADS) to improve the encoder's ability to characterize CT volumes. The cross-modal alignment is a more challenging pretext task that forces the encoder to learn better image representation ability. Deep self-distillation provides supervision to not only the bottleneck layer but also shallow layers, thus boosting the abilities of both. Comparative experiments show that, during pre-training, our CADS has lower computational complexity and GPU memory cost than competing SSL methods. Based on the pre-trained encoder, we construct PVT-UNet for 3D CT volume segmentation. Our results on seven downstream tasks indicate that PVT-UNet outperforms state-of-the-art SSL methods like MOCOv3 and DiRA, as well as prevalent medical image segmentation methods like nnUNet and CoTr. Code and pre-trained weight will be available at https://github.com/yeerwen/CADS.

Distribution shift widely exists in medical images acquired from different medical centres and poses a significant obstacle to deploying the pretrained semantic segmentation model in real-world applications. Test-time adaptation has proven its effectiveness in tackling the cross-domain distribution shift during inference. However, most existing methods achieve adaptation by updating the pretrained models, rendering them susceptible to error accumulation and catastrophic forgetting when encountering a series of distribution shifts (i.e., under the continual test-time adaptation setup). To overcome these challenges caused by updating the models, in this paper, we freeze the pretrained model and propose the Visual Prompt-based Test-Time Adaptation (VPTTA) method to train a specific prompt for each test image to align the statistics in the batch normalization layers. Specifically, we present the low-frequency prompt, which is lightweight with only a few parameters and can be effectively trained in a single iteration. To enhance prompt initialization, we equip VPTTA with a memory bank to benefit the current prompt from previous ones. Additionally, we design a warm-up mechanism, which mixes source and target statistics to construct warm-up statistics, thereby facilitating the training process. Extensive experiments demonstrate the superiority of our VPTTA over other state-of-the-art methods on two medical image segmentation benchmark tasks. The code and weights of pretrained source models are available at https://github.com/Chen-Ziyang/VPTTA.

Self-supervised learning (SSL) is an efficient pre-training method for medical image analysis. However, current research is mostly confined to certain modalities, consuming considerable time and resources without achieving universality across different modalities. A straightforward solution is combining all modality data for joint SSL, which poses practical challenges. Firstly, our experiments reveal conflicts in representation learning as the number of modalities increases. Secondly, multi-modal data collected in advance cannot cover all real-world scenarios. In this paper, we reconsider versatile SSL from the perspective of continual learning and propose MedCoSS, a continuous SSL approach for multi-modal medical data. Different from joint representation learning, MedCoSS assigns varying data modalities to separate training stages, creating a multi-stage pre-training process. We propose a rehearsal- based continual learning approach to manage modal conflicts and prevent catastrophic forgetting. Specifically, we use the k-means sampling to retain and rehearse previous modality data during new modality learning. Moreover, we apply feature distillation and intra-modal mixup on buffer data for knowledge retention, bypassing pretext tasks. We conduct experiments on a large-scale multi-modal unlabeled dataset, including clinical reports, X-rays, CT, MRI, and pathological images. Experimental results demonstrate MedCoSS's exceptional generalization ability across 9 downstream datasets and its significant scalability in inte- grating new modality data. The code and pre-trained model are available at https://github.com/yeerwen/MedCoSS.

How can we test AI performance? This question seems trivial, but it isn't. Standard benchmarks often have problems such as in-distribution and small-size test sets, oversimplified metrics, unfair comparisons, and short-term outcome pressure. As a consequence, good performance on standard benchmarks does not guarantee success in real-world scenarios. To address these problems, we present Touchstone, a large-scale collaborative segmentation benchmark of 9 types of abdominal organs. This benchmark is based on 5,195 training CT scans from 76 hospitals around the world and 5,903 testing CT scans from 11 additional hospitals. This diverse test set enhances the statistical significance of benchmark results and rigorously evaluates AI algorithms across various out-of-distribution scenarios. We invited 14 inventors of 19 AI algorithms to train their algorithms, while our team, as a third party, independently evaluated these algorithms on three test sets. In addition, we also evaluated pre-existing AI frameworks---which, differing from algorithms, are more flexible and can support different algorithms—including MONAI from NVIDIA, nnU-Net from DKFZ, and numerous other open-source frameworks. We are committed to expanding this benchmark to encourage more innovation of AI algorithms for the medical domain.

Cross-domain joint segmentation of optic disc and optic cup on fundus images is essential, yet challenging, for effective glaucoma screening. Although many unsupervised domain adaptation (UDA) methods have been proposed, these methods can hardly achieve complete domain alignment, leading to suboptimal performance. In this paper, we propose a triple-level alignment (TriLA) model to address this issue by aligning the source and target domains at the input level, feature level, and output level simultaneously. At the input level, a learnable Fourier domain adaptation (LFDA) module is developed to learn the cut-off frequency adaptively for frequency-domain translation. At the feature level, we disentangle the style and content features and align them in the corresponding feature spaces using consistency constraints. At the output level, we design a segmentation consistency constraint to emphasize the segmentation consistency across domains. The proposed model is trained on the RIGA+ dataset and widely evaluated on six different UDA scenarios. Our comprehensive results not only demonstrate that the proposed TriLA substantially outperforms other state-of-the-art UDA methods in joint segmentation of optic disc and optic cup, but also suggest the effectiveness of the triple-level alignment strategy.

Glaucoma is one of the leading causes of irreversible blindness. Segmentation of optic disc (OD) and optic cup (OC) on fundus images is a crucial step in glaucoma screening. Although many deep learning models have been constructed for this task, it remains challenging to train an OD/OC segmentation model that could be deployed successfully to different healthcare centers. The difficulties mainly comes from the domain shift issue, i.e., the fundus images collected at these centers usually vary greatly in the tone, contrast, and brightness. To address this issue, in this paper, we propose a novel unsupervised domain adaptation (UDA) method called Reconstruction-driven Dynamic Refinement Network (RDR-Net), where we employ a due-path segmentation backbone for simultaneous edge detection and region prediction and design three modules to alleviate the domain gap. The reconstruction alignment (RA) module uses a variational auto-encoder (VAE) to reconstruct the input image and thus boosts the image representation ability of the network in a self-supervised way. It also uses a style-consistency constraint to force the network to retain more domain-invariant information. The low-level feature refinement (LFR) module employs input-specific dynamic convolutions to suppress the domain-variant information in the obtained low-level features. The prediction-map alignment (PMA) module elaborates the entropy-driven adversarial learning to encourage the network to generate source-like boundaries and regions. We evaluated our RDR-Net against state-of-the-art solutions on four public fundus image datasets. Our results indicate that RDR-Net is superior to competing models in both segmentation performance and generalization ability.