PURPOSE: Stereotactic body radiation therapy (SBRT) for pancreatic cancer requires a skillful approach to deliver ablative doses to the tumor while limiting dose to the highly sensitive duodenum, stomach, and small bowel. Here, we develop knowledge-based artificial neural network dose models (ANN-DMs) to predict dose distributions that would be approved by experienced physicians. METHODS: Arc-based SBRT treatment plans for 43 pancreatic cancer patients were planned, delivering 30-33 Gy in five fractions. Treatments were overseen by one of two physicians with individual treatment approaches, with variations in prescribed dose, target volume delineation, and primary organs at risk. Using dose distributions calculated by a commercial treatment planning system (TPS), physician-approved treatment plans were used to train ANN-DMs that could predict physician-approved dose distributions based on a set of geometric parameters (vary from voxel to voxel) and plan parameters (constant across all voxels for a given patient). Patient datasets were randomly allocated, with two-thirds used for training, and one-third used for validation. Differences between TPS and ANN-DM dose distributions were used to evaluate model performance. ANN-DM design, including neural network structure and parameter choices, was evaluated to optimize dose model performance. RESULTS: Remarkable improvements in ANN-DM accuracy (i.e., from > 30% to < 5% mean absolute dose error, relative to the prescribed dose) were achieved by training separate dose models for the treatment style of each physician. Increased neural network complexity (i.e., more layers, more neurons per layer) did not improve dose model accuracy. Mean dose errors were less than 5% at all distances from the PTV, and mean absolute dose errors were on the order of 5%, but no more than 10%. Dose-volume histogram errors (in cm3 ) demonstrated good model performance above 25 Gy, but much larger errors were seen at lower doses. CONCLUSIONS: ANN-DM dose distributions showed excellent overall agreement with TPS dose distributions, and accuracy was substantially improved when each physician's treatment approach was taken into account by training their own dedicated models. In this manner, one could feasibly train ANN-DMs that could predict the dose distribution desired by a given physician for a given treatment site.
PURPOSE: Stereotactic body radiation therapy (SBRT) for pancreatic cancer requires a skillful approach to deliver ablative doses to the tumor while limiting dose to the highly sensitive duodenum, stomach, and small bowel. Here, we develop knowledge-based artificial neural network dose models (ANN-DMs) to predict dose distributions that would be approved by experienced physicians. METHODS: Arc-based SBRT treatment plans for 43 pancreatic cancerpatients were planned, delivering 30-33 Gy in five fractions. Treatments were overseen by one of two physicians with individual treatment approaches, with variations in prescribed dose, target volume delineation, and primary organs at risk. Using dose distributions calculated by a commercial treatment planning system (TPS), physician-approved treatment plans were used to train ANN-DMs that could predict physician-approved dose distributions based on a set of geometric parameters (vary from voxel to voxel) and plan parameters (constant across all voxels for a given patient). Patient datasets were randomly allocated, with two-thirds used for training, and one-third used for validation. Differences between TPS and ANN-DM dose distributions were used to evaluate model performance. ANN-DM design, including neural network structure and parameter choices, was evaluated to optimize dose model performance. RESULTS: Remarkable improvements in ANN-DM accuracy (i.e., from > 30% to < 5% mean absolute dose error, relative to the prescribed dose) were achieved by training separate dose models for the treatment style of each physician. Increased neural network complexity (i.e., more layers, more neurons per layer) did not improve dose model accuracy. Mean dose errors were less than 5% at all distances from the PTV, and mean absolute dose errors were on the order of 5%, but no more than 10%. Dose-volume histogram errors (in cm3 ) demonstrated good model performance above 25 Gy, but much larger errors were seen at lower doses. CONCLUSIONS:ANN-DM dose distributions showed excellent overall agreement with TPS dose distributions, and accuracy was substantially improved when each physician's treatment approach was taken into account by training their own dedicated models. In this manner, one could feasibly train ANN-DMs that could predict the dose distribution desired by a given physician for a given treatment site.
Authors: Ross A Abrams; Kathryn A Winter; William F Regine; Howard Safran; John P Hoffman; Robert Lustig; Andre A Konski; Al B Benson; John S Macdonald; Tyvin A Rich; Christopher G Willett Journal: Int J Radiat Oncol Biol Phys Date: 2011-02-01 Impact factor: 7.038
Authors: H M Ceha; G van Tienhoven; D J Gouma; C H Veenhof; C J Schneider; E A Rauws; S S Phoa; D González González Journal: Cancer Date: 2000-12-01 Impact factor: 6.860
Authors: Albert C Koong; Erin Christofferson; Quynh-Thu Le; Karyn A Goodman; Anthony Ho; Timothy Kuo; James M Ford; George A Fisher; Ralph Greco; Jeffrey Norton; George P Yang Journal: Int J Radiat Oncol Biol Phys Date: 2005-10-01 Impact factor: 7.038
Authors: Daniel T Chang; Devin Schellenberg; John Shen; Jeff Kim; Karyn A Goodman; George A Fisher; James M Ford; Terry Desser; Andrew Quon; Albert C Koong Journal: Cancer Date: 2009-02-01 Impact factor: 6.860
Authors: Michelle A Stinauer; Brian D Kavanagh; Tracey E Schefter; Rene Gonzalez; Thomas Flaig; Karl Lewis; William Robinson; Mark Chidel; Michael Glode; David Raben Journal: Radiat Oncol Date: 2011-04-08 Impact factor: 3.481
Authors: Elizabeth Huynh; Ahmed Hosny; Christian Guthier; Danielle S Bitterman; Steven F Petit; Daphne A Haas-Kogan; Benjamin Kann; Hugo J W L Aerts; Raymond H Mak Journal: Nat Rev Clin Oncol Date: 2020-08-25 Impact factor: 66.675