Intrafraction Motion of Patients During Cranial Treatments in Two Different Head Immobilization Systems

To determine the effectiveness of two different methods (reusable or patient specific) for head and neck immobilization during intracranial robotic radiosurgery treatments and infer the consequences of this motion for cone beam image guided treatments on conventional linacs. During intracranial robotic radiosurgery, the patient’s position is evaluated by acquiring a pair of orthogonal x-ray images of their skulls every 30-60 seconds. The robotic arm, on which the radiation treatment unit is mounted, corrects for patient position based on the last pair of images. For this system, the errors introduced by patient motion during treatment can be estimated from the difference between subsequent pairs of images. For most conventional linac treatments, the patient’s position is only determined at the start of treatment using 3D cone-beam imaging. For these systems, errors introduced by patient motion during treatment can also be estimated using the robotic system data, using the difference between the initial position and the average position over the typical duration of treatment (10 minutes). We examined the magnitude of patient motion during treatment for 59 patients using a reusable headrest and 36 patients using a disposable patient specific headrest. The positioning errors were evaluated using a well-known margin recipe which separates patient setup error into random, σ, and systematic, Σ, components. Results are summarised in table 1 for both headrests and both types of treatment modalities.Abstract 3738; Table 1Analysis of residual setup errors due to patient motion during treatment. Errors are estimated for the imaging systems used for the robotic radiosurgery system and for conventional radiation therapy treatments using a reusable headrest or a disposable but patient specific headrestHeadrestTranslational component (mm)Rotational component (degree)σΣMargin (mm)σΣRobotic Radiosurgery TreatmentsDisposable(Patient Specific)0.100.050.20.140.07Reusable0.290.180.60.260.15Conventional TreatmentsDisposable(Patient Specific)0.310.190.70.330.17Reusable0.460.261.00.340.17 Open table in a new tab For robotic radiosurgery treatments, random and systematic errors are approximately three to four times higher for the reusable headrest than for the patient specific system, stressing the importance of precise immobilization for this system, since no margin between the tumour and treated volume is typical. For conventional treatments, the relative differences between the two headrest systems are smaller, but relatively large compared the 3 mm margin currently used in our clinic for brain treatments. Care should be taken when considering changes to clinical practice which can impact the precision of radiation treatments.