FRAMELESS STEREOTAXY FOR PRETREATMENT PLANNING AND POSTTREATMENT EVALUATION OF RADIOSURGERY

In our centre, 111 patients have been treated with linear accelerator stereotactic radiosurgery. Angiographic, CT and MRI images are generat ed and the target coordinates calculated in 3 dimensions. For CT scann ing, cross sections of perpendicular and oblique fiducial markers are seen. For follow-up CT scans done without the frame, a virtual frame i s generated by means of a computer program that places fiducial marker s on each CT scan cut, as if the patient had been wearing the OBT fram e and the scan produced with the gantry parallel to the base of the fr ame. The position of the oblique marker may be calculated by knowing t he thickness and position of each CT cut. Various natural fiducial mar kers (bony landmarks) are identified by coordinates in the scan with t he patient wearing the real frame and in the scan with the virtual fra me applied. A transformation matrix is utilized to establish the equiv alence between the original CT scan with the real frame applied and su bsequent scans without the real frame but with the virtual frame appli ed. In effect, the virtual frame is re-applied in exactly the same pos ition as the real frame. Lesion measurements may then be duplicated an d growth or regression accurately established. The uncertainty in this system of re-application resides in possible patient movement, CT sca n slice thickness and inter-observer error in the identification of na tural fiducial markers.