Soft continuum manipulators integrating a shape sensing modality using optoelectronic sensors

在过去的几年中，已经开发出许多软连续体操纵器并使其适用于医疗用途，包括诸如微创手术（MIS）、内窥镜检查和结肠镜检查的手术。这些软连续体操纵器通常体积小且灵活，具有较大的工作空间，允许通过小通道和切口进入，以及通过身体安全地操纵，最大限度地减少手术中对软组织和器官的过度作用力。将这些操纵器用于医疗程序所面临的主要挑战之一是位置控制。重要的是要了解机械手在人体内的位置和形状，以确保使用的准确性和安全性。这是通过闭环位置反馈控制，使用不同的形状传感技术来完成的。视觉方法（摄像机/视频）非常简单，然而，在手术过程中，器官或物体可能会阻挡操纵器，导致摄像机无法确定位置。存在一种基于光纤布拉格光栅（FBG）的柔性连续体机械臂，它成功地集成了形状传感模态，其传感性能是准确的。尽管如此，其校准方法非常复杂。这种名为“传感器”的设备可以检测附着在软材料主体上的FBG的应变值，成本约为20，000英镑，并且制造形状传感器是复杂的。文献中的其他方法包括校准问题、漂移引起的误差、更新速度慢以及无法在更大范围的曲率上精确测量。为此，本项目提出了以下解决方案。这是一种新的形状传感方法的软连续体机械手，利用光电传感器。该解决方案解决了先前概述的问题，并且其具有潜在简化、准确、计算上便宜、易于集成以及具有低成本的优点，从而允许这种设备在医疗保健行业内容易地采用。这些传感器将被集成到一系列的圆盘，刚性连接，以形成柔性机械手的身体。它们根据连续表面上的反射量测量每个柔性机械结构中的距离（或变形）。根据测量的变形值，可以估计两到三个方向（滚动、俯仰和偏航）。因此，所有的估计方向在每个串联连接的机械弯曲可以简化的刚性连接模型，和整个机械手的形状可以可视化。该方法，形成本项目的一部分，首先包括机械手的结构设计，能够容纳这些传感器的数量。下一步将涉及开发软件，使传感器的数据测量和执行一组校准实验，以获得光电传感器的数据，物理取向值，使用各种建模和计算技术。下一步将是测试应用程序，通过将这种形状传感方法集成到具有选定驱动机构（肌腱，液压或气动）的机动致动器的柔性机械手中。这可以在幻影上进行测试，并且可以建立模拟以在三维中可视化操纵器形状。最后的目的是验证和设计一种关于如何将串联的内部刚性结构与由软材料组成的外罩结合的联合收割机设计。

Over previous years, many soft continuum manipulators have been developed and adapted for medical use, including procedures such as minimally invasive surgery (MIS), endoscopy and colonoscopy. These soft continuum manipulators are often small and flexible, with a large workspace, allowing access through small passages and incisions, as well as to safely manoeuvre through the body, minimising excessive force on soft tissue and organs in surgery. One of the main challenges that face the use of these manipulators for medical procedures is that of position control. It is important to have a sense of the position and shape of the manipulator inside the body, to ensure accuracy and safety in use. This is done through closed loop position feedback control, using different shape sensing techniques.A vision approach (camera/video) is very simple, however, during surgical procedures, organs or objects might block the manipulator, resulting the camera being unable to determine position. There exists a soft continuum manipulator successfully integrating a shape sensing modality based on Fibre Bragg gratings (FBG); their sensing performances are accurate. Despite this, its calibration method is very complicated. The device called "interrogator" which can detect strain values from the FBGs attached on the soft material body costs around £20,000 and manufacturing the shape sensor is complex. Other methods in the literature included issues with miniaturisation, errors due to drift, slow update rate, and inability to accurately measure over a larger range of curvatures.For this reason, the following solution is proposed for this project. This is a novel shape sensing method for soft continuum manipulators, that utilises optoelectronic sensors. This solution addresses the issues outlined previously, and it has the advantage of being potentially miniaturised, accurate, computationally inexpensive, easily integrated, as well as having a low cost, allowing such a device to be easily adopted within the healthcare industry. These sensors are to be integrated into a series of circular disks that are rigidly linked to form the body of the flexible manipulator. They measure the distances (or deformations) in each of the flexible mechanical structures, based in the amount of reflection on consecutive surfaces. From the measured deformation values, two to three orientations can be estimated (roll, pitch and yaw). Therefore, all of the estimated orientations in each of the serially connected mechanical flexures can be simplified by the rigid-link model, and the overall manipulator's shape can be visualised.The methods that form part of this project firstly include the structural design of the manipulator that is able to house a number of these sensors. Next will involve developing software that enables the measuring of data from the sensors and carrying out a set of calibration experiments to obtain the optoelectronic sensor's data to physical orientation values, using a various modelling and computational techniques. The next step will be testing the application, by integrating this shape sensing method in to a flexible manipulator with motorised actuators with a chosen driving mechanism (tendon, hydraulic, or pneumatic). This can be tested on phantoms and a simulation can be built to visualise the manipulator shape in three dimensions. The final aim is to miniaturise and devise a design on how to combine the serially connected internal rigid structure with with an outer cover consisting of soft material.