Semi-supervised learning (SSL) has greatly advanced 3D medical image segmentation by alleviating the need for intensive labeling by radiologists. While previous efforts focused on model-centric advancements, the emergence of foundational generalist models like the Segment Anything Model (SAM) is expected to reshape the SSL landscape. Although these generalists usually show performance gaps relative to previous specialists in medical imaging, they possess impressive zero-shot segmentation abilities with manual prompts. Thus, this capability could serve as "free lunch" for training specialists, offering future SSL a promising data-centric perspective, especially revolutionizing both pseudo and expert labeling strategies to enhance the data pool. In this regard, we propose the Generalist Model-driven Active Barely Supervised (GM-ABS) learning paradigm, for developing specialized 3D segmentation models under extremely limited (barely) annotation budgets, e.g., merely cross-labeling three slices per selected scan. In specific, building upon a basic mean-teacher SSL framework, GM-ABS modernizes the SSL paradigm with two key data-centric designs: (i) Specialist-generalist collaboration, where the in-training specialist leverages class-specific positional prompts derived from class prototypes to interact with the frozen class-agnostic generalist across multiple views to achieve noisy-yet-effective label augmentation. Then, the specialist robustly assimilates the augmented knowledge via noise-tolerant collaborative learning. (ii) Expert-model collaboration that promotes active cross-labeling with notably low labeling efforts. This design progressively furnishes the specialist with informative and efficient supervision via a human-in-the-loop manner, which in turn benefits the quality of class-specific prompts. Extensive experiments on three benchmark datasets highlight the promising performance of GM-ABS over recent SSL approaches under extremely constrained labeling resources.

Deep learning-driven 3D medical image segmentation generally necessitates dense voxel-wise annotations, which are expensive and labor-intensive to acquire. Cross-annotation, which labels only a few orthogonal slices per scan, has recently emerged as a cost-effective alternative that better preserves the shape and precise boundaries of the 3D object than traditional weak labeling methods such as bounding boxes and scribbles. However, learning from such sparse labels, referred to as barely-supervised learning (BSL), remains challenging due to less fine-grained object perception, less compact class features and inferior generalizability. To tackle these challenges and foster collaboration between model training and human expertise, we propose a Multi-Faceted ConSistency learning (MF-ConS) framework with a Diversity and Uncertainty Sampling-based Active Learning (DUS-AL) strategy, specifically designed for the active BSL scenario. This framework combines a cross-annotation BSL strategy, where only three orthogonal slices are labeled per scan, with an AL paradigm guided by DUS to direct human-in-the-loop annotation toward the most informative volumes under a fixed budget. Built upon a teacher-student architecture, MF-ConS integrates three complementary consistency regularization modules: (i) neighbor-informed object prediction consistency for advancing fine-grained object perception by encouraging the student model to infer complete segmentation from masked inputs; (ii) prototype-driven consistency, which enhances intra-class compactness and discriminativeness by aligning latent feature and decision spaces using fused prototypes; and (iii) stability constraint that promotes model robustness against input perturbations. Extensive experiments on three benchmark datasets demonstrate that MF-ConS (DUS-AL) consistently outperforms state-of-the-art methods under extremely limited annotation.

Medical image segmentation often requires large datasets with pixel-wise annotations. To alleviate this, we introduce BD-Net for semi-supervised and weakly supervised semantic segmentation (SWSS) using of sparse annotations (such as scribbles) alongside a few pixel-level annotations. BD-Net leverages a Boundary-Enhanced Module that improves boundary localization by extracting structural cues from annotated images through edge detection and multi-scale feature aggregation. Additionally, Density-Guided Contrastive Regularization enhances feature space compactness by using a contrastive loss informed by pixel density. Evaluation on the ACDC dataset shows BD-Net’s superior segmentation quality and generalizability compared to existing methods. The code will be available at https://github.com/Lemonzhoumeng/BD-Net.

Hand extension is crucial for stroke survivors with spasticity, where their fingers become rigid and their thumb remains curled within the palm. Due to the underactuated nature of the hand, the dominance of flexor muscles over extensors, and the limited surface area available, developing an extension glove with thumb assistance poses a challenge for researchers. This paper introduces a fully wearable soft hand extension glove based on the X-pouch and strap system, addressing the above challenges. The glove enables adequate finger extension, thumb abduction, and extension for high MAS score patients. Modelling and testing revealed extension torques of up to 2.7 Nm at the MCP joint and 0.67 Nm at the PIP and DIP joints. Performance evaluation, including comparison with existing methods, demonstrated the glove’s superior extension capabilities using a model hand with realistic stiffness. Furthermore, the glove’s effectiveness was confirmed through testing on a stroke patient with MAS = 2, validating its on-body functionality.

Segmenting prostate from magnetic resonance imaging (MRI) is a critical procedure in prostate cancer staging and treatment planning. Considering the nature of labeled data scarcity for medical images, semi-supervised learning (SSL) becomes an appealing solution since it can simultaneously exploit limited labeled data and a large amount of unlabeled data. However, SSL relies on the assumption that the unlabeled images are abundant, which may not be satisfied when the local institute has limited image collection capabilities. An intuitive solution is to seek support from other centers to enrich the unlabeled image pool. However, this further introduces data heterogeneity, which can impede SSL that works under identical data distribution with certain model assumptions. Aiming at this under-explored yet valuable scenario, in this work, we propose a separated collaborative learning (SCL) framework for semi-supervised prostate segmentation with multi-site unlabeled MRI data. Specifically, on top of the teacher-student framework, SCL exploits multi-site unlabeled data by: (i) Local learning, which advocates local distribution fitting, including the pseudo label learning that reinforces confirmation of low-entropy easy regions and the cyclic propagated real label learning that leverages class prototypes to regularize the distribution of intra-class features; (ii) External multi-site learning, which aims to robustly mine informative clues from external data, mainly including the local-support category mutual dependence learning, which takes the spirit that mutual information can effectively measure the amount of information shared by two variables even from different domains, and the stability learning under strong adversarial perturbations to enhance robustness to heterogeneity. Extensive experiments on prostate MRI data from six different clinical centers show that our method can effectively generalize SSL on multi-site unlabeled data and significantly outperform other semi-supervised segmentation methods. Besides, we validate the extensibility of our method on the multi-class cardiac MRI segmentation task with data from four different clinical centers.

Surface electromyography (sEMG) signal has been used in upper limb rehabilitation robots (ULRR). However, existing ULRR based on myoelectric controllers suffers from limited generalization ability in estimating three-dimensional (3-D) motion intention. This article proposes a muscle-synergy-inspired approach to enhance the generalization ability of the myoelectric controller of a cable-driven ULRR. Low-dimensional commands are extracted from sEMG signals based on an EMG-to-muscle activation model and non-negative matrix factorization. The extracted commands are used to estimate the 3-D human force. Two different trajectory tracking tasks are selected to test the generalization ability. The system is trained based on training sets where participants perform one task. Then the system is tested using testing sets where participants perform the other task. Finally, the system is verified on real-time robotic control experiment. Results show that the proposed controller achieves better force estimating accuracy, better trajectory tracking accuracy, and lower interaction force than the myoelectric controller without considering muscle synergies, which means the proposed controller yields better generalization performance.

Posture sensing of soft actuators is critical for performing closed-loop control of soft robots. This paper presents a novel end-to-end posture perception method for soft actuators by developing long short-term memory (LSTM) neural networks. A novel flexible bending sensor developed from off-the-shelf conductive silicon material was proposed and used for posture sensing. In the proposed method, the hysteresis of the soft robot and non-linear sensing signals from the flexible bending sensors have also been considered. With one-step calibration from the sensor output, the posture of the soft actuator could be captured by the LSTM network. The method was validated on a finger-size one DOF pneumatic fiber-reinforced bending actuator. Four kirigami-inspired flexible piezoresistive transducers were placed on the top surface of the actuator. Results show that the transducers could sense the posture of the actuator with acceptable accuracy. We believe our work could benefit soft robot dynamic posture perception and closed-loop control.

Semi-supervised learning has substantially advanced medical image segmentation since it alleviates the heavy burden of acquiring the costly expert-examined annotations. Especially, the consistency-based approaches have attracted more attention for their superior performance, wherein the real labels are only utilized to supervise their paired images via supervised loss while the unlabeled images are exploited by enforcing the perturbation-based "unsupervised" consistency without explicit guidance from those real labels. However, intuitively, the expert-examined real labels contain more reliable supervision signals. Observing this, we ask an unexplored but interesting question: can we exploit the unlabeled data via explicit real label supervision for semi-supervised training? To this end, we discard the previous perturbation-based consistency but absorb the essence of non-parametric prototype learning. Based on the prototypical networks, we then propose a novel cyclic prototype consistency learning (CPCL) framework, which is constructed by a labeled-to-unlabeled (L2U) prototypical forward process and an unlabeled-to-labeled (U2L) backward process. Such two processes synergistically enhance the segmentation network by encouraging morediscriminative and compact features. In this way, our framework turns previous "unsupervised" consistency into new "supervised" consistency, obtaining the "all-around real label supervision" property of our method. Extensive experiments on brain tumor segmentation from MRI and kidney segmentation from CT images show that our CPCL can effectively exploit the unlabeled data and outperform other state-of-the-art semi-supervised medical image segmentation methods.

Manually segmenting medical images is expertise-demanding, time-consuming and laborious. Acquiring massive high-quality labeled data from experts is often infeasible. Unfortunately, without sufficient high-quality pixel-level labels, the usual data-driven learning-based segmentation methods often struggle with deficient training. As a result, we are often forced to collect additional labeled data from multiple sources with varying label qualities. However, directly introducing additional data with low-quality noisy labels may mislead the network training and undesirably offset the efficacy provided by those high-quality labels. To address this issue, we propose a Mean-Teacher-assisted Confident Learning (MTCL) framework constructed by a teacher-student architecture and a label self-denoising process to robustly learn segmentation from a small set of high-quality labeled data and plentiful low-quality noisy labeled data. Particularly, such a synergistic framework is capable of simultaneously and robustly exploiting (i) the additional dark knowledge inside the images of low-quality labeled set via perturbation-based unsupervised consistency, and (ii) the productive information of their low-quality noisy labels via explicit label refinement. Comprehensive experiments on left atrium segmentation with simulated noisy labels and hepatic and retinal vessel segmentation with real-world noisy labels demonstrate the superior segmentation performance of our approach as well as its effectiveness on label denoising.

A weakly-supervised framework is proposed that cannot only make class inference but also provides reasonable decision basis in bone X-ray images. We implement it in three stages progressively: (1) design a classification network and use positive class activation map (PCAM) for attention location; (2) generate masks from attention maps and lead the model to make classification prediction from the activation areas; (3) label lesions in very few images and guide the model to learn simultaneously. We test the proposed method on a bone X-ray dataset. Results show that it achieves significant improvements in lesion location.

Soft robots attract research interests worldwide. However, its control remains challenging due to the difficulty in sensing and accurate modeling. In this paper, we propose a novel iterative learning model predictive control (ILMPC) method for soft bending actuators. The uniqueness of our approach is the ability to improve model accuracy gradually. In this method, a pseudo-rigid-body model is used to take an initial guess of the bending behavior of the actuator and the model accuracy is improved with iterative learning. Compared with conventional model free iterative learning control (ILC), the proposed method significantly reduces the learning curve. Compared with the model predictive control (MPC), the proposed method does not rely on an accurate model and it will output a satisfactory model after the learning process. A soft-elastic composite actuator (SECA) is used to validate the proposed method. Both simulation and experimental results show that the proposed method outperforms the conventional MPC and ILC.

A BCI-FES training platform has been designed for rehabilitation on chronic stroke patients to train their upper limb motor functions. The conventional functional electrical stimulation (FES) was driven by users’ intention through EEG signals to move their wrist and hand. Such active participation was expected to be important for motor rehabilitation according to motor relearning theory. The common spatial pattern (CSP) algorithm was applied as one pre-processing step in brain-computer interface (BCI) module to search for the optimal spatial projection direction after brain reorganization. The pre- and post- clinical assessment was conducted to identify the possible functional improvement after the training. Two chronic stroke subjects attended this pilot study and the error rate of the BCI control was less than 20% after training of 10 sessions. This implementation showed the feasibility for stroke patients to accomplish the BCI triggered FES rehabilitation training.