Lower-limb amputation not only impairs mobility but is also frequently accompanied by phantom limb pain and disrupted haptic and nociceptive perception in the residual limb, which can compromise prosthetic use and rehabilitation. At the residual limb–socket interface, excessive or uneven pressure may cause discomfort, inflammation, or even skin ulceration, making real-time localization of contact, pressure-intensity sensing, and timely detection of potentially harmful stimuli important for prosthetic fitting and safe use. Although flexible tactile sensors have improved substantially in sensitivity, stretchability, and stability, most systems still distinguish touch from pain using predefined thresholds and lack the intrinsic signal processing, memory, and synaptic plasticity of biological neural systems. “Recent approaches combining pressure sensors with neuromorphic elements such as memristors or transistors provide a route toward integrated sensing, processing, and memory. However, existing systems still struggle to simultaneously achieve precise real-time localization for haptic perception and the temporal and spatial integration required for pain sensing.” said the author Ye Qiu, a researcher at Zhejiang University of Technology, “Therefore, developing a bioinspired perceptual system capable of jointly decoding haptic location and intensity together with spatiotemporal pain information is an important step toward more effective sensory restoration and protective feedback in intelligent prostheses.”
This study developed a bioinspired perceptual sensor (BPS) with 2 synergistic pathways designed to emulate human haptic and pain perception. The haptic module used a P(VDF-TrFE) piezoelectric sensor to rapidly detect stimulus location, intensity, and application or release timing, while the pain module combined an LIG/PDMS piezoresistive sensor with a chitosan-gated ITO synaptic transistor, enabling synaptic-like short- and long-term plasticity and spatiotemporal integration of stimulus intensity, duration, frequency, and repetition. The 2 modules were integrated into a 2 × 2 bimodal flexible sensing array with signal-conditioning circuits and a microcontroller for parallel tactile sensing and neuromorphic pain processing. The researchers first characterized the sensitivity, response speed, stability, and plasticity-related behavior of the sensing components, and then integrated the BPS into a robotic hand to establish a closed-loop system capable of distinguishing harmless and harmful stimuli and triggering avoidance responses. Finally, the platform was deployed at the residual limb–prosthetic socket interface in participants with transtibial amputation, where haptic distribution and pain-warning signals were recorded during sitting, walking, stair climbing, jumping, and running to evaluate its feasibility for prosthetic fitting and rehabilitation feedback.
The results showed that the bioinspired perceptual sensor could simultaneously provide high-performance haptic detection and memory-like pain perception. The piezoresistive sensor achieved a sensitivity of 57.9 kPa⁻¹ in the low-pressure range, with response and recovery times of 31 and 43 ms, while the piezoelectric sensor reached response and recovery times as fast as 5 and 4 ms and remained stable after 2,000 loading cycles. When integrated with the synaptic transistor, the pain pathway exhibited EPSC, paired-pulse facilitation, and transitions from short- to long-term plasticity, while accumulating responses according to stimulus intensity, duration, frequency, and repetition to emulate pain sensitization and desensitization. In closed-loop robotic experiments, the system distinguished harmless from harmful stimuli, triggered rapid avoidance once the pain threshold was exceeded, and lowered its subsequent warning threshold after a painful experience, producing a conditioned-reflex-like protective response. In feasibility trials with participants with transtibial amputation, the BPS simultaneously monitored pressure distribution and pain-related risk at the residual limb–socket interface during sitting, walking, stair climbing, jumping, and running, helping identify abnormal loading and support movement adjustment. These findings demonstrate its potential for prosthetic fitting, residual-limb protection, and rehabilitation training.
The significance of this work lies in integrating real-time haptic localization with spatiotemporal pain accumulation within a single bioinspired perceptual system, providing sensory feedback for intelligent prostheses that more closely resembles human somatosensory processing. By combining a piezoelectric haptic pathway with a neuromorphic pain pathway, the BPS can not only identify stimulus location and intensity but also emulate pain-related features such as accumulation, sensitization, and desensitization, supporting closed-loop robotic avoidance and prosthetic rehabilitation feedback. Compared with conventional pressure-mapping or fixed-threshold warning systems, this approach provides more comprehensive synergistic haptic and pain information, with potential applications in prosthetic fit assessment, residual-limb health monitoring, and movement adjustment. “However, currently human trials are mainly for feasibility verification. In the future, we will expand the number of subjects and types of amputations, and further improve wireless integration, medical grade packaging, long-term wearing stability, and sensor fixation reliability during dynamic motion, thereby promoting the development of this system towards long-term wearable intelligent prosthetics and clinical rehabilitation applications.” said Ye Qiu.
Authors of the paper include Ye Qiu, Shihan Wang, Qiangqiang Qian, Yu Yan, Weisheng Wang, Xiu Jia, Shengwei Fan, Yuan Bao, Ye Tian, Yi Song, Aiping Liu, Liu Wang, Liqiang Zhu, and Huaping Wu.
This work was supported by the National Natural Science Foundation of China (Grant nos. U25A20294, 12572184, 12372168, 12388101, 12272351, and 12532008), the National Key Research and Development Program of China (2024YFB3816500), the Natural Science Foundation of Zhejiang Province of China (LRG25A020001 and LZ24A020004), and the Fundamental Research Funds for the Provincial Universities of Zhejiang (RF-A2025017).
The paper, “A Bioinspired Perceptual Sensor for Spatiotemporal Decoding of Haptic and Pain Stimuli in Prostheses” was published in the journal Cyborg and Bionic Systems on Sept 26, 2026, at https://doi.org/10.34133/cbsystems.0679.
Cyborg and Bionic Systems