Polyp segmentation is vital for early colorectal cancer detection, yet traditional fully supervised methods struggle with morphological variability and domain shifts, requiring frequent retraining. Additionally, reliance on large-scale annotations is a major bottleneck due to the time-consuming and error-prone nature of polyp boundary labeling. Recently, vision foundation models like Segment Anything Model (SAM) have demonstrated strong generalizability and fine-grained boundary detection with sparse prompts, effectively addressing key polyp segmentation challenges. However, SAM's prompt-dependent nature limits automation in medical applications, since manually inputting prompts for each image is labor-intensive and time-consuming. We propose OP-SAM, a One-shot Polyp segmentation framework based on SAM that automatically generates prompts from a single annotated image, ensuring accurate and generalizable segmentation without additional annotation burdens. Our method introduces Correlation-based Prior Generation (CPG) for semantic label transfer and Scale-cascaded Prior Fusion (SPF) to adapt to polyp size variations as well as filter out noisy transfers. Instead of dumping all prompts at once, we devise Euclidean Prompt Evolution (EPE) for iterative prompt refinement, progressively enhancing segmentation quality. Extensive evaluations across five datasets validate OP-SAM's effectiveness. Notably, on Kvasir, it achieves 76.93% IoU, surpassing the state-of-the-art by 11.44%.

Tracheotomy is commonly performed for patients needing prolonged intubation, airway obstruction, and neck injuries. Accurate placement of the incision and the tracheal window is paramount in order to avoid complications. Current surgical technique heavily relies on palpating cartilage landmarks on the neck to place the incision. In order to achieve the accelerated goals of the robot-assisted subtask in a tracheotomy, this paper proposes a novel autonomous palpation-based acquisition strategy - RASEC in the tracheal region, which can interactively predict the next acquisition point to maximize the expected information and minimize the costs of palpation procedure. We employ a Gaussian Process (GP) to model the distribution of hardness and utilize anatomical information as a priori input to guide the point of palpation for medical robots. The dynamic tactile sensor based on the resonant frequency is introduced to measure tissue hardness in the tracheal region by millimeter-scale contact to secure the interaction. We investigate the kernel fusion method to blend the Squared Exponential (SE) kernel with the Ornstein-Uhlenbeck (OU) kernel and optimize the Bayesian optimization search by leveraging the anatomical information of the larynx as a priori knowledge. Moreover, we further regularize the exploration and greed factors. The tactile sensor’s moving distance and the robotic base link’s rotation angle during the incision localization process are considered new factors in the acquisition strategy. Simulation and physical phantom experiments are conducted for comparison with state-of-the-art GP-based exploration approaches. The results show that the sensor’s moving distance was reduced by 53.1% and the rotation angle of the base was reduced by 75.2% of the previous values without sacrificing overall performance capabilities. The satisfying algorithmic index (average precision 0.932, average recall 0.973, average F1 score 0.952) with fewer central estimation distance errors (0.423 mm) and high resolution (1 mm) indicates the performance of the proposed RASEC in terms of exploration efficiency, cost awareness, and localization accuracy for incision localization and recommendation in real robot-assisted subtask in the tracheotomy procedure.Note to Practitioners—This work is well motivated to introduce the Level of Autonomy (LoA) 2 - task-level autonomy, specifically in the context of tracheotomy procedures. The incorporation of robotic palpation techniques aims to provide surgeons with enhanced capabilities for incision recommendations, which directly benefit surgeons to visualize hands-on information and localize the trachea regions more efficiently and further reduce cognitive load. To detect the trachea region for intubation incision without costly ergodic acquisition, this article suggests a highly efficient acquisition strategy utilizing the fusion kernel function and regularized impact factors, eliminating the time consumption for such localization task. The actual clinical value is that our proposed strategy can earn more time for further increasing the probability of patient resuscitation, to facilitate supervised autonomy in the real clinic scene.

Deep learning has demonstrated exceptional performance in medical image analysis, but its effectiveness degrades significantly when applied to different medical centers due to domain shifts. Lesion detection, a critical task in medical imaging, is particularly impacted by this challenge due to the diversity and complexity of lesions, which can arise from different organs, diseases, imaging devices, and other factors. While collecting data and labels from target domains is a feasible solution, annotating medical images is often tedious, expensive, and requires professionals. To address this problem, we combine active learning with domain-invariant feature learning. We propose a Domain-shift Active Learning (DistAL) framework, which includes a transferable feature learning algorithm and a hybrid sample selection strategy. Feature learning incorporates contrastive-consistency training to learn discriminative and domain-invariant features. The sample selection strategy is called RUDY, which jointly considers Representativeness, Uncertainty, and DiversitY. Its goal is to select samples from the unlabeled target domain for cost-effective annotation. It first selects representative samples to deal with domain shift, as well as uncertain ones to improve class separability, and then leverages K-means++ initialization to remove redundant candidates to achieve diversity. We evaluate our method for the task of lesion detection. By selecting only 1.7% samples from the target domain to annotate, DistAL achieves comparable performance to the method trained with all target labels. It outperforms other AL methods in five experiments on eight datasets collected from different hospitals, using different imaging protocols, annotation conventions, and etiologies.

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.

Precise image segmentation provides clinical study with instructive information. Despite the remarkable progress achieved in medical image segmentation, there is still an absence of a 3D foundation segmentation model that can segment a wide range of anatomical categories with easy user interaction. In this paper, we propose a 3D foundation segmentation model, named SegVol, supporting universal and interactive volumetric medical image segmentation. By scaling up training data to 90K unlabeled Computed Tomography (CT) volumes and 6K labeled CT volumes, this foundation model supports the segmentation of over 200 anatomical categories using semantic and spatial prompts. To facilitate efficient and precise inference on volumetric images, we design a zoom-out-zoom-in mechanism. Extensive experiments on 22 anatomical segmentation tasks verify that SegVol outperforms the competitors in 19 tasks, with improvements up to 37.24\% compared to the runner-up methods. We demonstrate the effectiveness and importance of specific designs by ablation study. We expect this foundation model can promote the development of volumetric medical image analysis. The model and code are publicly available at https://github.com/BAAI-DCAI/SegVol.