面向脊柱外科切削手术的机器人触听觉术野仿生认知

In order to prevent possible damages to the important tissues caused by high-speed rotating tool, the spinal surgical robot need recognize the surgical field (an isolated area where surgery is performed). The content to be recognized includes discriminating the tissues in the surgical field, estimating the relative position between tool and the tissues, and monitoring the weak disturbance to the tool during cutting operation. The project aims to simulate the mechanisms of surgeon's tactile and auditory processing, and the surgical field recognition of the spinal surgical robot can be realized through acquiring, processing and identifying the tactile and auditory information, so the accurate operation can be performed by the robot. The main research contents include four aspects. Firstly, inspired by frequency and spatial pattern discrimination ability of human vibrotactile, we will investigate the surgical field recognition based on robot tactile ability. Secondly, inspired by the bandpass filtering in human cochlea and nerve signal processing in the auditory cortex, we will investigate the feature extraction from the sound pressure signal and the surgical field recognition. Thirdly, inspired by the multisensory integration exhibited by the sensory cortex, we will investigate the identification of weak disturbance based on the bio-inspired integration of auditory and tactile of the robot. Finally, the surgical field recognition method will be experimentally verified through the laminectomy in porcine spines.

为了防止高速旋转的刀具对重要组织的损伤，在脊柱外科切削手术中需要机器人对术野(手术时的操作范围)进行认知，认知的内容包括识别术野中的组织，估计术野中刀具与组织的相对位置关系，以及监测术野中刀具的微弱扰动。本项目模拟手术医生的触听觉信息处理机制，通过获取、处理和识别脊柱手术机器人的触听觉信息，实现术野认知，从而能够完成精细的手术操作。具体研究内容包括四个方面：模拟人振动触觉频率与空间模式分辨的神经机制，研究基于触觉仿生的机器人术野认知；模拟耳蜗对声压信号的带通滤波机制和人脑听觉中枢对神经冲动的处理方法，研究基于听觉仿生的声压信号特征提取和术野认知；模拟人脑感觉皮层对多种感觉信息的整合机制，研究基于机器人触听觉信息仿生融合的微弱扰动辨识；最后通过在猪脊柱上进行的椎板切除手术验证提出的术野认知方法的有效性。

使用手术机器人可以提高脊柱椎板减压等常见骨科手术的安全性、稳定性和精准性。为了防止高速旋转的刀具对重要组织的损伤，在脊柱外科切削手术中需要机器人对术野(手术时的操作范围)进行认知，认知的内容包括识别术野中的组织，估计术野中刀具与组织的相对位置关系，以及监测术野中刀具的微弱扰动。本项目模拟手术医生的触听觉信息处理机制，通过获取、处理和识别脊柱手术机器人的触听觉信息，实现术野认知，从而能够完成精细的手术操作。具体研究内容包括四个方面：模拟人振动触觉频率与空间模式分辨的神经机制，研究了基于触觉仿生的机器人术野认知；模拟耳蜗对声压信号的带通滤波机制和人脑听觉中枢对神经冲动的处理方法，研究了基于听觉仿生的声压信号特征提取和术野认知；模拟人脑感觉皮层对多种感觉信息的整合机制，研究了基于机器人触听觉信息仿生融合的微弱扰动辨识。本项目还进行了基于振动触觉的骨科手术机器人骨曲面铣削研究，利用实验室搭建的手术机器人分析骨曲面铣削过程中的不同方向上的振动信号变化趋势，并通过实验得到的特征控制机器人完成了曲面铣削过程。