Description
Stereotactic Radiosurgery is a commonly used treatment process for treating cranial lesions. This thesis addresses an alternative to the current process that utilizes a CT simulation for the development of a treatment plan. Because it is now common to combine an MRI with the CT in order to better define the tumor margins, this work proposes eliminating the CT simulation altogether and making better use of the cone-beam CT (CBCT). The CBCT modality is currently being used in conjunction with the simulation CT as a repositioning tool. Instead of using it to just reposition the patient for the treatment; we will use it to establish the treatment positions as well as to verify a MRI pre-plan. This process will eliminate the treatment repositioning error inherent in the necessity to reproduce the planning position. This process will also increase patient convenience by shortening the amount of time the patient needs to be at the facility and potentially lead to treating the patient the same day as the consultation. Three areas needed to be investigated for this project to be completed. The first was to ensure that similar or better dose accuracy could be achieved between the CBCT-based treatment plan and the CT-based treatment plan. The second obstacle to address was the development of a semi-immobilization device that would be quick to produce and simple to use. Finally, we needed to determine an appropriate way to QA the delivery of the verified CBCT plan without removing the patient from the treatment couch. The first goal was accomplished by exporting and anonymizing (in order to ensure patient confidentiality) data from 5 patients. This data included MR, CBCT, and CT images as well as the original treatment plan based on the CT image. We used the Eclipse treatment planning system, (Varian Medical Systems, Palo Alto, CA), to optimize the existing plan on the MR image and establish the optimized plan as the pre-plan for the patient. The re-optimized plan was then verified and approved on the CBCT image, eliminating the need for the CT simulation. These plans were imported into IMSure (Standard Imaging, Middleton, WI), a software that compares the dose calculated in the treatment planning system with a simplified dose calculation to a point in a homogeneous medium. The second goal addressed the complication of saving time in the production of a semi-immobilization device that was both comfortable and effective. Two different devices were tested for their ability to immobilize volunteers. Each volunteer was monitored with AlignRT (VisionRT, London, UK), a surface imaging system, with varying results. The first device was a universal, flexible blue headmold; one size fits all. The second device was a choice of 5 hard plastic neck supports currently being used for CT scans. The final goal, the QA of the plan, comes into play because we wish to leave the patient on the treatment couch in the same position as when he/she comes in for the CBCT in order to eliminate the repositioning error. Three steps will be used to ensure the plan will be delivered accurately; the first being running the plan through portal dosimetry on an alternate linear accelerator, the second, comparing the first segments of each field on the treatment screen to the calculated segments in Eclipse, and the final consists of comparing the MR preplan Modulation Complexity Score, MCS, with the verified CBCT re-optimized plan. Our results affirmed the re-optimized CBCT plan to be a viable treatment option for simple cranial lesions. The dose accuracy was shown to be comparable to the current CT method. The immobilization devices still require more research before they can be used, if they can be used at all. The current FMB immobilization device, a frame-less, mask-less, and biteblock-less individualized headcast, is an acceptable alternative, though it takes more time to make. It has been proven to be successful at immobilizing compliant patients. Lastly, the delivery QA precautions should alert the physicist if there would be any complications in the delivery of the treatment plans. After more time and a larger pool of data, the MCS score could prove to be a valuable QA option, even on its own.
