The TALON removable head frame system for stereotactic radiosurgery/radiotherapy: Measurement of the repositioning accuracy

Purpose: To present the TALON removable head frame system as an immobilizat ion device for single-fraction intensity-modulated stereotactic radiosurger y (IMRS) and fractionated stereotactic intensity-modulated radiotherapy (FS -IMRT); and to evaluate the repositioning accuracy by measurement of anatom ic landmark coordinates in repeated computed tomography (CT) examinations. Methods and Materials: Nine patients treated by fractionated stereotactic i ntensity-modulated radiotherapy underwent repeated CTs during their treatme nt courses. We evaluated anatomic landmark coordinates in a total of 26 rep eat CT data sets and respective x, y, and z shifts relative to their positi ons in the nine treatment-planning reference CTs. An iterative optimization algorithm was employed using a root mean square scoring function to determ ine the best-fit orientation of subsequent sets of anatomic landmark measur ements relative to the original image set. This allowed for the calculation or the x, y, and z components of translation of the target isocenter for e ach repeat CT. In addition to absolute target isocenter translation, the ma gnitude (sum vector) of isocenter motion and the patient/target rotation ab out the three principal axes were calculated. Results: Anatomic landmark analysis over a treatment course of 6 weeks reve aled a mean target isocenter translation of 0.95 +/- 0.55, 0.58 +/- 0.46, a nd 0.51 +/- 0.38 mm in x, y, and z directions, respectively. The mean magni tude of isocenter translation was 1.38 +/- 0.48 mm. The 95% confidence inte rval ([CI], mean translation plus two standard deviations) for repeated iso center setup accuracy over the 6-week period was 2.34 mm. Average rotations about the x, y. and z axes were 0.41 +/- 0.36, 0.29 +/- 0.25, and 0.18 +/- 0.15 degrees, respectively. Analysis of the accuracy of the first repeated setup control, representative of single-fraction stereotactic radiosurgery situations, resulted in a mean target isocenter translation in the x, y, a nd z directions of 0.52 +/- 0.38, 0.56 +/- 0.30, and 0.46 +/- 0.25 mm, resp ectively. The mean magnitude of isocenter translation was 0.99 +/- 0.28 mm. The 95% confidence interval for these radiosurgery situations was 1.55 mm. Average rotations at first repeated setup control about the x, y, and z ax es were 0.24 +/- 0.19, 0.19 +/- 0.17, and 0.19 +/- 0.12 degrees, respective ly. Conclusion: The TALON relocatable head frame was seen to be well suited for immobilization and repositioning of single fraction stereotactic radiosurg ery treatments. Because of its unique removable design, the system was also seen to provide excellent repeat immobilization and alignment for fraction ated stereotactic applications. The exceptional accuracy for the single-fra ction stereotactic radiosurgical application of the system was seen to dete riorate only slightly over a 6-week fractionated stereotactic treatment cou rse. (C) 2001 Elsevier Science Inc.