The goal of this proposal is to improve the potential of dose escalation studies in the lung and prostate by delivering dose more accurately and precisely. Emphasis will be on using novel 3D imaging techniques to improve tumor identification for treatment planning, and to reduce organ motion and patient setup uncertainty, principal factors that limit the accuracy of treatment in the lung and prostate. In the pelvis, rectal complications limit the dose that can be delivered to the prostate, while variability in rectal and bladder filling causes changes in prostate position between treatments. We shall develop methods and computerized tools to investigate the use of magnetic resonance spectroscopic imaging as a guide for ?dose painting? with intensity-modulated radiation treatment (IMRT), to selectively intensify the radiation dose to certain tumor-bearing regions within the prostate. Further, 50 patients will be entered into a study in which the target position relative to the radiation field is corrected each fraction, using a CT scanner housed in the treatment room to acquire images immediately prior to treatment. In the treatment of non small cell lung cancer, respiration-associated tumor motion is a severe limitation. Our plan is to combine respiratory gating (RG) of the linac with IMRT, and develop image-guided methods to further improve accuracy. Specifically, we shall correlate spiral CT images with respiration to achieve """"""""4-D"""""""" imaging, i.e., provide spatial and temporal information of the tumor volume, for the planning of RG-IMRT. These data will be analyzed to identify the respiration phase of minimum tumor motion for the specific patient, determine the residual tumor motion within the treatment gate interval to define appropriate treatment apertures, and compare treatment plans at different phases. In addition, we shall implement megavoltage conebeam CT imaging (MVCBI) in 50 patients, using the therapeutic beam and an amorphous silicon electronic portal imaging device, to visualize the tumor during the initial several RG treatments, correct for any errors in tumor position and, if necessary, adjust treatment apertures. The effectiveness of the above correction strategies in lung and prostate will be assessed relative to alternative strategies and to the current practice of portal image-based correction, using dosimetric indices calculated from the serial image data sets. Common to all the studies is the investigation of computer algorithms to determine changes in organ shape between image data sets. We will examine three applications: first, to transfer MR-detected regions of tumor in prostate to the planning CT scan; second, to automatically localize organs in CT and MVCB images; and third, to calculate cumulative dose to organs from multi-fraction treatment. The validity of these tools will be tested in phantom and patient studies.
Showing the most recent 10 out of 121 publications
