A positioning sensor device for biomedical robot : development and application to a laser skin stimulator

Since the dawn of time, humans have tried to reduce pain by various means and to understand its origin. Today, tremendous advances in science and technology have shown that pain is not a simple concept. In most of the cases, minor pain can be treated or eased by a physiotherapist or by using different kinds of drugs. However, more complex feelings like chronic pain are difficult to treat. Studying how the nervous system of humans work and how the brain interprets the pain signals is still a challenge. Among the people working for the Institute of NeuroSciences (IoNS) of the Université catholique de Louvain (UCL) \nomenclature[A]{UCL}{Université catholique de Louvain}, the neurologist and professor André Mouraux carries out research on pain. To perform his experiments, he is currently using a laser stimulator to briefly heat the hand or foot dorsum so that the person feels a minor pain. This information is sent to the brain thanks to sensors sensitive to pain (nociceptors) through the peripheral nervous system. In some experiments, this operation is repeated approximately 40 times. The goal is to obtain constant stimuli as much as possible. Therefore, before every stimulation, the laser must be positioned precisely with respect to the skin surface : at a distance varying from 85 mm to 115 mm and perpendicular to it with a maximal error of 15°. Since the experiment is tedious and tiring, André Mouraux is looking for a system allowing to hold and correctly position the laser with respect to the skin. The list of specifications expresses that two separated devices are needed to fulfill the whole need. On the one hand, a robot must hold and move the laser in space. On the other hand, a sensing device needs to give information about the skin surface. This information is then sent to the robot as a movement to place the laser correctly compared to the skin. This master thesis is precisely dedicated to this sensing device. The objective is to design, build and test such a system. Firstly, information about points belonging to the skin surface needs to be known. For this purpose, two technologies are identified : 3D scanners and distance sensors. A state of the art of existing technologies is presented. From these researches, the technologies are discussed and two possible solutions are chosen : a 3D scanner using structured light technology and a system composed of three distance sensors using laser triangulation. Since neither of these methods can be differentiated on paper, they are implemented and compared based on predefined criteria. An experimental protocol is followed with the help of an industrial robot to hold and position precisely both systems in space. It allows to come to the conclusion that both methods fulfill the requirements. On the one hand, the 3D scanner is accurate since the maximal angular error observed is of 6°. However, it is not able to take into account possible movements of the subject. If the patient moves, a new scan taking approximately 75 seconds is necessary to update the scene. On the other hand, the system with three distance sensors is less accurate since the angular error sometimes exceed the 15° limit. Nevertheless, it happens in a few complex situations. Its main advantage is that it is able to update its measurements 5 times per second making it capable of taking into account the possible patient movement. Finally, conclusions are drawn on the work performed during this final year and possible ways of improvements are proposed for a potential subsequent work.