Zhuang Fu

Objective. Multi-sweep freehand 3D ultrasound is affected by two types of trajectory degradation: inter-sweep rigid bias and intra-sweep local tracking jitter, both of which hinder high-fidelity volumetric reconstruction from tracked 2D ultrasound slices. Although recent neural radiance field methods enable joint optimization of pose and neural representation in optical imaging, they are not directly transferable to ultrasound since limited slice overlap and strong view-dependent speckle of ultrasound make photometric supervision alone ill posed. Approach. This paper presents GLA-NeRF, a global-local aligned neural radiance field framework for joint optimization of pose and neural representation in multi-sweep freehand ultrasound. Within this framework, a unified pose model decouples sweep-wise global rigid bias from frame-wise local tracking jitter, allowing all slices to be mapped into a common canonical space. A global registration module is introduced to correct inter-sweep misalignment through appearance-based image retrieval, hierarchical matching, geometric inlier verification, and uncertainty-weighted Mahalanobis distance. A local denoising module is further incorporated to suppress intra-sweep tracking jitter through statistical sensor noise priors and cumulative Lie-group B-splines, which provide variance-reduced pseudo-labels and physically plausible kinematic constraints. Main results. Experiments on in vivo ultrasound datasets demonstrate that GLA-NeRF stabilizes joint optimization, reduces tracking errors at both global ( 3.95 mm/ 1.53 ∘ accuracy regardless of initialization) and local (from 2.5 mm/ 2.0 ∘ to 0.65 mm/ 0.52 ∘ ) scales, and improves the anatomical fidelity of synthesized ultrasound views. Significance. Joint optimization of pose and implicit neural representations allow the reconstructed representation to become both more geometrically coherent and more stable for ultrasound novel view synthesis.

This paper presents the design, modeling, and control of a heavy-duty scissor lift platform intended to meet the demands of high-intensity operations in specialized work environments. We completed the structural design of the lifting platform and implemented the calibration and optimization using ANSYS Finite Element Analysis (FEA). The kinematic and dynamic models are elaborated. Then we validated the dynamic model with Adams. The results show that the maximum error of the proposed model does not exceed 2%. Leveraging the validated dynamics, a feed-forward controller paired with a PD feedback loop is designed. Simulation tests show a position error that is less than 10 cm, and no observable overshoot is found when tracking a 1.0-meter step. A cross-coupling synchronous controller is also developed and validated. The proposed framework simultaneously enhances structural safety and motion precision for heavy-duty lift systems.

Background: In recent years, artificial intelligence (AI) technology has experienced significant growth, leading to the development of advanced tools that assist radiologists in image interpretation and diagnostic decision-making. In the field of medical image processing, object detection and segmentation are crucial research areas. Achieving automatic and rapid segmentation of organs and lesions can significantly enhance physicians’ work efficiency. However, most existing networks feature complex architectures and entail high computational complexity. Due to the limited processing power of the devices, achieving real-time segmentation often proves challenging. Therefore, the aim of this study was to design a lightweight segmentation algorithm that reduces resource consumption and enhances real-time performance. Methods: Firstly, we propose a compact U-shaped network called the weak redundancy U-Net (WRU-Net) specifically designed for real-time segmentation tasks in medical imaging. By reducing the number of channels, feature redundancy across all scales is reduced, compelling the network to utilize resources efficiently at every level. Furthermore, we propose “auxiliary information flows” to facilitate the propagation of phased results, thereby enhancing the decoder. Secondly, this paper introduces a novel visual prompt mode that differs from the standard object detection mode. We refer to this as “object prompt”, which means visualizing the position of the target object in an image to guide the viewer. Unlike standard object detection tasks that provide bounding boxes, this paper achieves the aforementioned effect in the form of heatmaps. Correspondingly, we propose a network for heatmap prediction, which further simplifies task complexity and achieves semantic detection of the target object. Results: We conducted experiments using the datasets of the thyroid nodule, chest, placental vessel, brain tumors, heart, liver, and spleen, which include medical images of X-ray, computed tomography (CT), and ultrasound. Furthermore, we conducted comparative experiments using several state-of-the-art networks. The segmentation performance of the networks was evaluated using metrics such as the Dice similarity coefficient (DSC), intersection over union (IoU), precision, and recall. The average DSC, IoU, precision, and recall of our model across each dataset were 88.53%, 82.58%, 86.34%, and 84.85%, respectively. Regarding efficiency, our WRU-Net achieved 130.67 KB model size and 488 frames per second (FPS), outperforming larger models. For heatmap prediction, our network exhibited a similarly efficient profile with a parameter size of only 24.75 KB and a speed of 9,862.13 FPS on graphics processing unit (GPU) and 107.93 FPS on central processing unit (CPU). Conclusions: In scenarios where the calculation ability of the equipment is constrained or real-time performance requirements are stringent, the design of lightweight networks serves as a fundamental approach to achieving high-speed data stream processing functionality. For instance, this applies to embedded devices and real-time ultrasonic scanning applications. The method proposed in this paper significantly reduces the number of parameters, thereby decreasing computational resource consumption and enhancing speed. This improvement holds substantial significance for optimizing computational efficiency in practical applications.

In this paper, a fault-tolerant control (FTC) scheme based on reinforcement learning (RL) is proposed to address the optimal FTC challenge in multi-actuator aircraft system with uncertain parameters and time-varying state constraints. The control algorithm consists of two parts, i.e., the state constraints observer based on collision detection trigger and the optimal FTC policy based on actor-critic architecture. The state constraints are caused by the actuator blockage fault. To estimate these constraints, an adaptive external torque observer and a collision detection algorithm are devised. By integrating state constraints estimation with a barrier function, a multi-actuator cooperative FTC policy based on RL is formulated. Ultimately, the efficacy of the proposed scheme is demonstrated through experimental validation tests.

Flexible needle percutaneous puncture demands precise and efficient path planning to ensure surgical safety and success. However, traditional algorithms often struggle to balance real-time performance and planning quality in high-resolution, complex three-dimensional environments. This paper introduces a Neural-Guided RRT^* (NG-RRT^*) algorithm to address these challenges. First, we propose a 3D U-Net based entry points selection method, leveraging spatial features to generate the probability distributions of optimal entry points. Second, we develop a neural-guided non-uniform sampling strategy which utilizes an optimal path prediction network to improve the sampling efficiency of RRT^*. Additionally, we introduce a native arc trajectory search strategy that directly generates kinematically feasible curve segments, reducing path complexity and improving smoothness. Simulation results demonstrate that the proposed NG-RRT^* outperforms previous methods in path quality, computational efficiency, and adaptability to complex scenarios, and exhibits good clinical feasibility in actual anatomical environments. This work provides a robust and efficient solution for flexible needle path planning, advancing the field of medical navigation systems.