Zhuoqi Cheng

Electrical bioimpedance (EBI) measurement provides insights into the biophysical properties of tissues, offering valuable information for tumor diagnosis and classification. Deep learning has demonstrated distinct advantages in analyzing complex biomedical data. However, their applications in the rapid diagnosis of brain tumors had not been fully explored. In this study, 52 brain tumor samples were collected for EBI measurement. A deep learning framework that integrates multi-scale (MS) impedance feature extraction with frequency band attention was developed for the analysis of bioimpedance spectra (1–349 kHz) and automatic tumor classification. The model used parallel convolutional kernels (sizes 1, 3, 5, 7, 9) to capture local and global features, alongside an attention module to prioritize diagnostic frequency bands. Model performance was evaluated using precision, sensitivity, specificity and F 1-score. Significant differences in impedance values were observed among gliomas, meningiomas, and metastases. The proposed model exhibits high sensitivity and precision in tumor classification tasks, achieving F 1-scores of 91.54% (gliomas vs meningiomas vs metastases), 99.61% (glioma vs metastasis), 93.12% (lower-grade gliomas vs glioblastomas), and 98.75% (1p/19q codeleted vs non-codeleted gliomas), with significant conductivity differences ( p < 0.05) between tumor types. In summary, the proposed framework, which integrates MS features and adaptive frequency, improves the performance of EBI-based tumor classification, and shows promise as an accurate intraoperative tool for the rapid diagnosis of brain tumors.

Background: Accurate navigation within the lungs is critical for bronchoscopy procedures but remains challenging due to the complex airway anatomy. Artificial intelligence (AI) integrating real-time landmark detection and navigation guidance may enhance precision and safety. Methods: We developed an AI Bronchial detection and Airway Navigation system (AIBAN) for bronchoscopy, combining airway lumen detection, anatomic localization, and navigation guidance. A YOLOv8-based detection module was used to identify 34 airway landmarks in real time. Next, an airway-level localization module estimated the bronchoscope’s position using a hierarchical anatomic graph, identifying the deepest branch compatible with all coherent detections in each frame. A navigation module then recommended the next airway to follow along predefined anatomic paths, providing directional cues to perform a complete and systematic bronchoscopy. Results: AIBAN was developed on a data set of 18 bronchoscopy videos, using 15 for training and 3 for validation. In validation, the framework correctly identified the bronchoscope’s location in 91.1% of the frames. In addition, 92.6% of predictions were within one airway branch of the true location, and 94.5% were within 2 branches. The navigation guidance performance was also promising, with the system following the correct anatomic pathway in 96.2% of test videos, demonstrating reliable guidance through the lungs. Conclusion: AIBAN enables accurate airway detection, localization, and navigation guidance in a simulation-based setting. It can provide reliable guidance along predefined paths and precise spatial positioning and is a promising tool for medical training and has potential for future clinical implementation.

The article presents the results of a manipulability analysis and optimal remote center of motion (RCM) frame placement based on the manipulability index for a simplified model of a planar laparoscopic surgical robot system with a control-based RCM constraint. The paper presents an analytical kinematic model of the planar laparoscopic surgical robot system, a method for computing a manipulability objective taking into account the RCM constraint, and a formulation of the optimization problem used to establish the optimal placement of the system with respect to the task. The presented results are backed by a sensitivity analysis measuring the influence of parameter uncertainty on the value of the objective.

Needle-based interventions are routine in clinical practice, yet clinicians lack reliable, precise, real-time estimates of the tip-to-target distance—–a key determinant of safety and efficiency. To bridge this critical clinical gap, physics-based models that maps electrical bio-impedance (EBI) measurements to the absolute tip-to-target distance by leveraging tissue-intrinsic electrical properties encoded in EBI are presented in this work. A tripolar electrode configuration is adopted to ensure tool-independence and broad compatibility across instruments and anatomical sites. The optimal layout of electrode radius, inter-electrode spacing, and placement was identified through a systematic parametric study. The model was validated across finite-element simulations, saline solution experiments, and ex vivo layered tissue. In all settings, proximity distance estimates were accurate and robust across different insertion angles. In simulation, it achieved 0.55 mm RMSE/1.06 mm MAE when the target layer was less conductive than the first layer and 0.40 mm RMSE/1.21 mm MAE when it was more conductive. In saline, four insertion angles yielded millimeter errors with no angle dependence. In ex vivo layered tissue, accuracy remained near the millimeter level across angles (RMSE 1.00–1.13 mm and MAE 1.30–1.58 mm), which is slightly lower than saline yet consistent across orientations. Together, these results establish a general, instrument-agnostic framework for absolute proximity distance sensing during needle interventions with the potential to enhance safety and efficiency in minimally invasive procedures. Furthermore, by first focusing on a two-layer model, this work provides a clear pathway to multi-layer tissue modeling and broader anatomical generalization.

The integration of robotics and mobile networks (5G/6G) through the Internet of Robotic Things (IoRT) is revolutionizing telemedicine, enabling remote physician participation in scenarios where specialists are scarce, where there is a high risk to them, such as in conflicts or natural disasters, or where access to a medical facility is not possible. Nevertheless, touching a human safely with a robotic arm in non-engineered or even out-of-hospital environments presents substantial challenges. This article presents a novel IoRT approach for healthcare in or from remote areas, enabling interaction between a specialist’s hand and a robotic hand. We introduce the IoRT-in-hand: a smart, lightweight end-effector that extends the specialist’s hand, integrating a medical instrument, an RGB camera with servos, a force/torque sensor, and a mini-PC with Internet connectivity. Additionally, we propose an open-source Android app combining MQTT and ROS for real-time remote manipulation, alongside an Edge–Cloud architecture that links the physical robot with its Digital Twin (DT), enabling precise control and 3D visual feedback of the robot’s environment. A proof of concept is presented for the proposed tele-robotic system, using a 6-DOF manipulator with the IoRT-in-hand to perform an ultrasound scan. Teleoperation was conducted over 2300 km via a 5G NSA network on the operator side and a wired network in a laboratory on the robot side. Performance was assessed through human subject feedback, sensory data, and latency measurements, demonstrating the system’s potential for remote healthcare and emergency applications. The source code and CAD models of the IoRT-in-hand prototype are publicly available in an open-access repository to encourage reproducibility and facilitate further developments in robotic telemedicine.

Vascular diseases are commonly diagnosed using Ultrasound (US) imaging, which can be inconsistent due to its high dependence on the operator’s skill. Among these, Deep Vein Thrombosis (DVT) is a common yet potentially fatal condition, often leading to critical complications like pulmonary embolism. Robotic US Systems (RUSs) aim to improve diagnostic test consistency but face challenges with the complex scanning pattern requiring precise control over US probe pressure, such as the one needed for indirectly detecting occlusions during DVT assessment. This work introduces an imitation learning method based on Kernelized Movement Primitives (KMP) to standardize the contact force profile during US exams by training a robotic controller using sonographer demonstrations. A new recording device design enhances demonstration acquisition, integrating with US probes and enabling seamless force and position data recording. KMPs are used to link scan trajectory and interaction force, enabling generalization beyond the demonstrations. Our approach, evaluated on synthetic models and volunteers, shows that the KMP-based RUS can replicate an expert’s force control and US image quality, even under conditions requiring compression during scanning. It outperforms previous methods using manually defined force profiles, improving exam standardization and reducing reliance on specialized sonographers.

Needle intervention is widely employed in clinical practices such as biopsies, regional anesthesia, blood sampling, neurosurgery, and brachytherapy. Traditional needle insertion relies on surgeon expertise and kinesthetic feedback, yet accurately targeting deep tissue structures remains challenging. To address this, significant research has advanced sensing technologies to aid insertion accuracy. This paper comprehensively reviews recent developments in needle insertion sensing techniques, encompassing needle tip position tracking, proximity measurement, and puncture detection. It evaluates these methods across metrics including accuracy, cost-effectiveness, portability, compatibility, noise resistance capability, Technology Readiness Level (TRL), and future trends. Emerging research directions highlight advancements in machine learning integration, miniaturization, and enhanced multimodal sensing capabilities to improve procedural outcomes and expand application domains.

Deep needle insertion to a target often poses a huge challenge, requiring a combination of specialized skills, assistive technology, and extensive training. One of the frequently encountered medical scenarios demanding such expertise includes the needle insertion into a femoral vessel in the groin. Conventionally, deep needle insertion is guided by ultrasound imaging. However, utilizing ultrasound for needle tracking demands specialized training and skill to manipulate the probe effectively and interpret the imaging accurately. To address this challenge, this paper presents an innovative technology for needle tip real-time tracking. This advancement will be instrumental in facilitating robotic-guided needle insertion toward the identified target. Specifically, our approach revolves around the creation of scattering imaging using an optical fiber-equipped needle, and uses Convolutional Neural Network (CNN) based algorithms to enable real-time estimation of the needle tip’s position and orientation during insertion procedures. The efficacy of the proposed technology was rigorously evaluated through three experiments. The first two experiments involved rubber and bacon phantoms to simulate groin anatomy. The positional errors averaging 2.3±1.5 mm and 2.0±1.2 mm, and the orientation errors averaging 0.2±0.11 rad and 0.16±0.1 rad. Furthermore, the system’s capabilities were validated through experiments conducted on fresh porcine phantom mimicking more complex anatomical structures, yielding positional accuracy results of 3.2±3.1 mm and orientational accuracy of 0.19±0.1 rad. Given the average femoral arterial radius of 4 to 5 mm, the proposed system is demonstrated with a great potential for precise needle guidance in femoral artery insertion procedures. In addition, the findings highlight the broader potential applications of the system in the medical field.