Autosegmentation of prostate anatomy for radiation treatment planning using deep decision forests of radiomic features
- Meghan W Macomber ,
- Mark Phillips ,
- Ivan Tarapov ,
- Rajesh Jena ,
- Aditya Nori ,
- David Carter ,
- Loic Le Folgoc ,
- Antonio Criminisi ,
- Matthew J Nyflot
Physics in Medicine & Biology | , Vol 63(23)
Machine learning for image segmentation could provide expedited clinic workflow and better standardization of contour delineation. We evaluated a new model using deep decision forests of image features in order to contour pelvic anatomy on treatment planning CTs.