Room: Karl Dean Ballroom C
Purpose: Current image guided prostate radiotherapy often relies on the use of implanted fiducial markers (FMs) or transducers for online/offline target localization. Fiducial or transducer insertion requires an invasive procedure that adds cost and risks for bleeding, infection and discomfort to the patient. We are developing a novel markerless prostate localization strategy using a pre-trained deep learning model to interpret routine projection kV X-ray images without the need for daily cone-beam computed tomography (CBCT).
Methods: A deep learning model was first trained by using several thousand annotated projection X-ray images. The end point of the deep learning was a model capable of identifying the location of the prostate target for a given input X-ray projection image. To assess the accuracy of the approach, four patients with prostate cancer received volumetric modulated arc therapy (VMAT) were retrospectively studied. The results obtained by using the deep learning model and the actual positon of the prostate were compared quantitatively.
Results: After training, the model identified the prostate location on a projection kV image within 200 ms. The deviations between the target positions obtained by the deep learning model and the corresponding annotations ranged from 0.7 mm to 2.21 mm for anterior-posterior (AP) direction, and from 1.23 mm to 2.20 mm for lateral direction. Target position provided by deep learning model for the kV images acquired using OBI is found to be consistent that derived from the implanted FMs.
Conclusion: This study demonstrates that highly accurate markerless prostate localization based on deep learning is achievable. The strategy provides a clinically valuable solution for image guided radiotherapy of the prostate and should be adaptable to other soft tissue tumor targets. The proposed method may be useful for daily patient positioning and real-time target tracking during prostate radiotherapy.