The paper introduces Dual Co-Train, a source‑free domain adaptation framework for ultrasound tongue segmentation that operates under extreme data scarcity. Starting from a lightweight UltraUNet pretrained on only five labeled images, the method iteratively refines pseudo‑labels, filters unreliable masks with a contour‑based quality‑control module, and generates target‑style synthetic image‑mask pairs via a segmentation‑guided conditional GAN. The student model is trained on a mix of clean pseudo‑labeled target images, noisy pseudo‑labels with consistency regularization, and synthetic samples, enabling closed‑loop adaptation without access to source data. Experiments on 12 source‑target transfer pairs across eight datasets show that this approach improves segmentation overlap and contour accuracy over both unsupervised baselines and supervised models.
arXiv:2608.22532v1 Announce Type: new Abstract: Domain shift across imaging modalities and acquisition sites remains a significant barrier to the clinical deployment of segmentation models. Source-fr...
The paper introduces Segment Anything Small (SAS), a data‑augmentation method that improves deep‑learning segmentation of small anatomical structures in ultrasound images. SAS uses two transformations: resizing and embedding organ thumbnails into a black background to vary organ scale, and adding noise to regions of interest to mimic tissue texture variability. Experiments on one internal and five external datasets show Dice score gains up to 0.35, with an average improvement of 0.16, and demonstrate that SAS enhances model robustness and generalizability without adding hallucinations or artifacts.
The paper introduces ExiL, a mask‑conditioned progressive learning framework for bone ultrasound segmentation that models annotation as a structured refinement trajectory. ExiL uses a synthetic expert‑like brush simulator and a lightweight U‑Net to learn from imperfect masks, and it can be updated in real time from expert refinements. In experiments on UltraBones100k and a prospective volunteer dataset, ExiL cut average annotation time from 60 to 20 seconds per frame and improved mean Dice by about 0.045, achieving 0.87 Dice and 2.7 px boundary error with 10–50 ms inference.
arXiv:2609.39681v1 Announce Type: new Abstract: Unsupervised domain adaptation (UDA) reduces the annotation burden in panoptic segmentation by leveraging a cost-effectively labeled source domain (e.g...
US-JEPA introduces a self‑supervised framework for ultrasound imaging that predicts masked latent representations instead of raw pixels, using a frozen, domain‑specific teacher to provide stable targets. This approach avoids the hyperparameter sensitivity and computational cost of traditional online teachers, enabling the student model to build upon the teacher’s semantic priors. The authors benchmark US‑JEPA against all publicly available ultrasound foundation models on UltraBench, showing competitive or superior performance across multiple organs and pathological conditions under linear probing.
The paper introduces LowBridge, a method for cross‑modal medical image segmentation that leverages shared low‑level features such as edges between MRI and CT scans. It trains a generative model to reconstruct source‑modality images from edge maps and then trains a segmentation network on these generated images. At test time, edge features from target‑modality images are fed into the generative model to produce source‑style images, which are segmented by the pretrained network, achieving state‑of‑the‑art results on multiple public datasets.
arXiv:2605. 25402v2 Announce Type: replace-cross Abstract: Self-supervised pre-training paradigm has gained increasing prominence for learning transferable representations in medical imaging, yet existing methods for ultrasound (US) images operate at the image or frame level, overlooking the anatomical context for clinical-aligned representation learning.
Match4Annotate is a test‑time framework that transfers user‑specified annotations from a labeled ultrasound video to an unlabeled target video without requiring manual initialization. It uses a spatiotemporal implicit feature representation, a continuous implicit feature flow for alignment, and flow‑guided annotation transfer to unify sparse point and dense mask transfer. The method achieves state‑of‑the‑art performance on four clinical ultrasound datasets, outperforming dense feature‑matching baselines and one‑shot segmentation methods, and works without task‑specific training on a single consumer GPU.
arXiv:2606. 29586v1 Announce Type: cross Abstract: Vision-language foundation models have shown strong potential in medical image analysis.
The paper introduces ANT, a test‑time adaptation framework that improves prostate cancer detection in micro‑ultrasound by performing a segmentation‑guided adaptation. ANT aligns a pretrained detection encoder to the target domain’s prostate anatomy using pseudo‑masks from a frozen segmentation network, thereby correcting domain‑specific feature drift while preserving cancer‑discriminative features. In a leave‑one‑center‑out evaluation, ANT raises mean AUC by 2.9% at the biopsy‑core level and 3.6% at the patient level compared to no adaptation, outperforming existing TTA baselines.
InstEditSeg is a generative framework that treats medical segmentation as an instruction-driven image editing task. Instead of producing binary masks, it renders a color-coded overlay on the original image guided by textual instructions, leveraging latent diffusion models to align with natural image distributions and reduce domain gaps. The method incorporates a DINOv3 visual encoder and a multi-scale feature pyramid fused into the diffusion U‑Net, and uses a dual‑branch classifier‑free guidance strategy to lower inference cost, achieving competitive accuracy on polyp and skin lesion datasets while improving cross‑domain generalization and multi‑lesion segmentation.
The paper introduces an unsupervised domain adaptation framework that aligns redundancy-reducing features to enable accurate 3D segmentation of cone-beam CT (CBCT) without target-domain annotations or inference-time adaptation. The method is architecture-agnostic, working with both CNN-based and ViT-based foundation models, and is evaluated on two liver segmentation benchmarks for interventional vascular procedures and radiation therapy. Results show that even large pretrained segmentation networks need explicit feature-space bridging to generalize across diagnostic CT and CBCT, and the proposed approach consistently outperforms existing pretrained foundation models and UDA strategies.