Basic Study for Constructing Training/Simulation System for Medical Operations

我们从多个方面研究了构建医疗操作模拟/训练系统。首先，我们引入了弹丸模型来模拟弹性体的行为，这是模拟医疗手术所必需的。提出了一种新的具有最小长度约束的弹簧模型。开发的模型不仅可以变形，而且可以解剖和切割。接下来，我们专注于技能训练。我们提出了一种有效的运动技能训练方法，即从教师的角度出发，利用振动线索和教师的视线，给出期望的动作，如运动摆动。我们构建了一个一维系统，并对提出的训练方法进行了评估。实验结果表明，教师的视力是有效的。提出了一种利用视觉和加速度计在头盔显示器(HMD)上精确叠加图像的方法。利用加速度信息预测头部运动，补偿端到端系统延迟，使基于视觉的跟踪具有较强的鲁棒性。实验结果表明，即使用户快速移动头部，该方法也能准确地叠加合成图像。最后，研究了模拟多个虚拟物体的碰碰型触觉设备的路径规划问题。碰碰式方法让我们摆脱了佩戴或手持触觉设备的限制，获得了真正的触觉。该方法特别适用于医疗手术等数量有限的已知形状物体的仿真，其研究结果可为未来医疗手术仿真/训练系统的开发奠定基础。

We have studied from various aspects toward constructing a simulation/training system for medical operations. First, we have introduced spring-dumpier model to simulate the behavior of elastic objects, which is necessary to simulate medical operations. We have proposed a new spring model that has minimum length constraint. The developed model allows us not only to deform but also to dissect and cut.Next, we focused on skill training. We have proposed an efficient training method for motor skills, such as sport swing motions, by using vibration cue and teacher's sight, i.e., giving the desired motion from the teacher's viewpoint,. We constructed a one-dimensional system and evaluated the proposed training method. The experimental result shows that the teacher's sight is effective. The vibration cue is effective only when the teacher's sight is withdrawn.Then, we have proposed a method a method for accurate image overlay on head-mounted displays (HMDs) using vision and accelerometers. Acceleration information is used for predicting the head motion to compensate the end-to-end system delay and to make the vision-based tracking robust. Experimental results showed that the proposed method can overlay synthetic images accurately even if the user moves his head very quickly. This technique is essential to realize "What You See IS What You Feel" or WYSIWYF situation, which is important for training systems.Finally, we have studied path planning of encountered-type haptic device that simulates multiple virtual objects. Encountered-type approach allows us to be free from the constraints of worn or held haptic device and to get real sense of touch. This approach is effective especially when we simulate limited number of objects with known shape like medical operations.The obtained results in this research could be foundations toward simulation/training systems for medical operations in the future.