Purpose: Current high-dose-rate (HDR) prostate brachytherapy catheter placement is based on physicians’ experience. It may lead to large plan quality variations among physicians, and increased placement randomness for individual physician. In this study, we proposed a learning-based method to predict HDR needle position for prostate cancer patients.
Methods: The proposed learning-based framework consists of deformable registration network (Reg-Net), multi-atlas ranking and catheter regression. Binary masks of the organs-at-risks are first transformed into distance maps which describe the distance of each local voxel to the organ surfaces. For a new patient, Reg-Net deformably registers the distance maps from multi-atlas set to match this patient’s distance map and then bring catheter maps from multi-atlas to this patient. Based on several similarity criterions, we select the top-ranked atlas’ deformed catheter positions as predicted catheter position for this patient. Finally, catheter regression is used to refine the final catheter positions. A retrospective study on 90 patients with a five-fold cross-validation was used to evaluate the proposed method’s feasibility. Plans were optimized based on both the clinical and predicted catheter pattern.
Results: For all patients, on average, both clinical plan dose and new plan dose meet the common dose constraints when prostate dose coverage is kept at V100=95%. The plans from predicted catheter pattern have slightly higher hotspot in terms of V150 by 5.0% and V200 by 2.9% on average. For bladder V75, rectum V75 and urethra V125, the average difference is close to zero, and the range of most patients is within ±1cc.
Conclusion: We developed a novel learning-based method in predicting catheter locations for HDR prostate brachytherapy. It has great clinical potential since it can provide catheter location estimation prior to catheter placement, which could reduce the dependence on physicians’ experience in catheter implantation and improve the quality of prostate HDR treatment plans.