知能機械情報学とコンピュータ外科による脳外科への展開

Computer-assisted neurosurgery enhanced by mechano-informatics

　はじめに

　定位脳手術を軸として，工学分野と脳神経外科分野は互いに寄与しあう側面が多い。神経計測，画像診断，放射線治療とならんで，工学技術が脳神経外科手術に深く寄与してきたものには，ニューロナビゲーション，手術ロボット，術中画像誘導手術が挙げられる。本稿の目的は，これらの分野について工学的解説を加えた後，現在工学分野で発展している技術のうち，特に脳神経外科手術に将来役立ちそうな内容の紹介を行うことである。特に，工学支援で可能となった新しい核磁気共鳴画像（magnetic resonance imaging：MRI）誘導手術，手術ロボット，最後にナビゲーション手術用立体ディスプレイ等について述べる。

　Image-guided interventions based on intra-operative medical images are promising method of cure to achieve minimal invasiveness. Furthermore the image-guidance integrated with robotic system has provided great assistance for surgeons to execute the planned surgical maneuver in precise manner. However, the preoperative images often used in image-guided robotic surgery are not suitable in certain applications where inevitable internal movements due to the respiration, heartbeat, or insertion pain occur. In such cases, the mis-alignment between the preoperatively prepared plan on the image and the patient in robotic surgery could cause unexpected complications.

　The objective of the paper is to report state-of-art and future perspective on real-time image-guided medical robot. The proposed method can compensate for the motion and guide the surgery in real-time by using intra-operative images as a guiding tool, which is otherwise impossible by related surgical robot guided by pre-operative images.

　First robot to be presented is designed for use in ultrasound-image guided needle insertion surgery. A novel ultrasonic image segmentation technique was also proposed and applied to support real-time operation. A visual servo control was employed to involve the extracted image features in the robot operation. The robot has the one-unit structure composed of an ultrasound probe and a needle delivery unit to maintain the constant relationship in position between the probe and needle, thus ensuring the needle observed in ultrasonic images.

　The second robot is MRI-compatible robot where more detailed anatomical and/or physiological image observation are possible compared to ultrasound imager. The challenge in achieving in MRI compatibility includes but not limited to material selection and over all design. We also fast motion detection technique integrated in MR imaging, using MR image projection profile as detection target.

　In summary, the intra-operative image-driven robot will have potential impact on surgery, in which more accurate and less-invasive surgery is possible by motion compensation combined with robotic surgery.