| 要約: | For the first time in 1985, robots were introduced to the neurosurgical operating rooms. Since then this field of robotics have advanced with new technologies with better accuracy and safety. However, developing robots for neurosurgery is a challenging task due to the sensitive nature of these applications and additionally their high cost of development. A typical neurosurgical robot costs millions of dollars in 2018, even though the costs of components and calculations have been lowered tremendously. This is mainly due to focus of current neurosurgical robots on few highly precise and specific surgical tasks which makes their use limited. In this thesis, development of a neurosurgical robot with a focus on more general surgical tasks related to surgical navigation is investigated, and its design and development are reported. Using this platform to develop more general surgical applications brings down the cost. Furthermore, this system has been developed using open source platforms that continually updates new features and technologies, helping this platform to stay up to date with current science of the day. To have a unified approach, three key phases had to be included. The first was the data collection and analysis, which involved collecting actual surgical data from live surgeries to analyze the workspace of the targeted surgical tasks which helped in structural design of the robot. The second was the design and development of the actual robot based on the initial analysis of the surgeries. In this work, the robot was designed to be small and dexterously suited to neurosurgical operating rooms. The third phase was the assessment of the robot’s function. Using any new device in actual surgeries requires a long track record of fault free operations. Therefore, the robot had to be validated using cadavers or animal subjects however, these methods also bring ethical, logistics and cost issues. So, a new approach was designed for assessment of the robot. This method involves using rapid prototyping techniques and actual patient data. In the process of this work, robotic applications had to be developed from scratch to enable the robot to perform surgical tasks and applications such as semi-automated patient-image registration, biopsy needle guidance and endoscope manipulation. Patient-image registration function matches the physical coordinates of the patient with medical image coordinates autonomously. Furthermore, biopsy needle guidance for brain tissue sampling is a common feature in neurosurgical robots, which enables a suitable benchmark for a robot’s accuracy. Endoscope manipulation is also a topic of interest in neurosurgery as it introduces different challenges compared to Laparoscopic surgeries. The robot performed with adequate accuracy for general surgical tasks, however, it showed some limitations in the endoscope manipulation features.
|
|---|
