A verification method of image registration for image-guided therapy (IGT)

Image-guided therapy (IGT) provides an image data (CT, MRI and PET) to overcome a deficiency of sight during operation by using augmented reality thus tracking performance is important. However, it is difficult to verify the tracking performance in medical applications because makers which can detect the pose in image registration cannot be attached to the patient´s bodies. This paper proposes a method to verify the tracking performance of image registration method for IGT and shows one verification example. An experiment is designed to measure tracking performance of image registration method. A 3D model is manufactured based on CT images and commercial haptic device is utilized to overcome difficulties of measuring accuracy. The experimental result is analyzed in terms of mean and standard deviation and verified by using the just-noticeable-difference (JND). The proposed verification procedure provides a guideline of the criterion of tracking accuracy to a therapy which doesn´t have any quantified reference for IGT.