Biomimetics

This study aimed to investigate the timing of foot-off and initial contact at the end of the first walking training session with a Wearable Power-Assist Locomotor (WPAL) in novice healthy users. Eight healthy volunteers with no walking experience with the WPAL participated in this study. The participants walked back and forth on a straight 5 m path for 60 min with the WPAL. We calculated the differences between the participant’s foot-off and initial contact timing, as well as the start and end timing of the pre-programmed WPAL lower-limb swing time. Data were divided into four segments of 100 data points. We calculated the median of the last 100 data points and examined whether it falls within an appropriate time range. The foot-off timing tended to be within the appropriate time range (median, −0.44 s); however, the initial contact timing was earlier than the appropriate time range (median, −0.17 s). Although some participants performed foot-off within the appropriate time range, all performed initial contact earlier than the appropriate time range. These findings may contribute to establishing practice protocols for stable walking with wearable robotic exoskeletons in patients with spinal cord injury.

Highly sensitive bioinspired cutaneous receptors are essential for realistic human-robot interaction. This study presents a biomimetic tactile sensor morphologically modeled after the Meissner corpuscle, designed for high dynamic sensitivity achieved using a coiled configuration. Our proposed electrolytic polymerization technique with magnet-responsive hybrid fluid (HF) was employed to fabricate soft, elastic rubber sensors with embedded coiled electrodes. The coiled configuration, optimized by electrolytic polymerization, exhibited high responsiveness to dynamic motions including pressing, pinching, twisting, bending, and shearing. The mechanism of the haptic property was analyzed by electrochemical impedance spectroscopy (EIS), revealing that reactance variations define an equivalent electric circuit (EEC) whose resistance (R_p), capacitance (C_p), and inductance (L_p) change with applied force; these changes correspond to mechanical deformation and the resulting variation in the sensor’s built-in voltage. The roll-type Meissner-inspired sensor demonstrated fast-adapting behavior and broadband vibratory sensitivity, indicating its potential for high-performance tactile and auditory sensing. These findings confirm the feasibility of electrolytically polymerized hybrid fluid rubber as a platform for next-generation bioinspired haptic interfaces.

Artificial intelligence (AI) is increasingly shaping interventional oncology, with growing interest in its application across thermal ablation modalities such as radiofrequency ablation (RFA), cryoablation, high-intensity focused ultrasound (HIFU), and microwave ablation (MWA). This review characterises the current landscape of AI-enhanced thermal ablation, with particular emphasis on emerging opportunities within MWA technologies. We examine how AI-driven methods—convolutional neural networks, radiomics, and reinforcement learning—are being applied to optimise patient selection, automate image segmentation, predict treatment response, and support real-time procedural guidance. Comparative insights are provided across ablation modalities to contextualise the unique challenges and opportunities presented by microwave systems. Emphasis is placed on integrating AI into clinical workflows, ensuring safety, improving consistency, and advancing personalised therapy. Tables summarising AI methods and applications, a conceptual workflow figure, and a research gap analysis for MWA are included to guide future work. While existing applications remain largely investigational, the convergence of AI with advanced imaging and energy delivery holds significant promise for precision oncology. We conclude with a roadmap for research and clinical translation, highlighting the need for prospective validation, regulatory clarity, and interdisciplinary collaboration to support the adoption of AI-enabled thermal ablation into routine practice.