Mattea L Welch1, Chris McIntosh2, Benjamin Haibe-Kains3, Michael F Milosevic4, Leonard Wee5, Andre Dekker5, Shao Hui Huang6, Thomas G Purdie7, Brian O'Sullivan6, Hugo J W L Aerts8, David A Jaffray9. 1. Department of Medical Biophysics, University of Toronto, Canada; The Techna Institute for the Advancement of Technology for Health, Toronto, Canada; Princess Margaret Cancer Centre, University Health Network, Toronto, Canada. 2. Radiation Medicine Program, Princess Margaret Cancer Centre, Toronto, Canada; The Techna Institute for the Advancement of Technology for Health, Toronto, Canada; Princess Margaret Cancer Centre, University Health Network, Toronto, Canada. 3. Department of Medical Biophysics, University of Toronto, Canada; Ontario Institute of Cancer Research, Toronto, Canada; Princess Margaret Cancer Centre, University Health Network, Toronto, Canada; Vector Institute, Toronto, Canada. 4. Department of Radiation Oncology, University of Toronto, Canada; Radiation Medicine Program, Princess Margaret Cancer Centre, Toronto, Canada; Princess Margaret Cancer Centre, University Health Network, Toronto, Canada. 5. Department of Radiation Oncology (MAASTRO), GROW Research Institute, Maastricht University, The Netherlands. 6. Department of Radiation Oncology, University of Toronto, Canada; Princess Margaret Cancer Centre, University Health Network, Toronto, Canada. 7. Department of Radiation Oncology, University of Toronto, Canada; Radiation Medicine Program, Princess Margaret Cancer Centre, Toronto, Canada; The Techna Institute for the Advancement of Technology for Health, Toronto, Canada; Princess Margaret Cancer Centre, University Health Network, Toronto, Canada. 8. Dana-Farber Cancer Institute, Brigham and Women's Hospital, Harvard Medical School, Boston, USA. 9. Department of Medical Biophysics, University of Toronto, Canada; Department of Radiation Oncology, University of Toronto, Canada; IBBME, University of Toronto, Canada; Radiation Medicine Program, Princess Margaret Cancer Centre, Toronto, Canada; The Techna Institute for the Advancement of Technology for Health, Toronto, Canada; Princess Margaret Cancer Centre, University Health Network, Toronto, Canada. Electronic address: david.jaffray@rmp.uhn.ca.
Abstract
PURPOSE: Refinement of radiomic results and methodologies is required to ensure progression of the field. In this work, we establish a set of safeguards designed to improve and support current radiomic methodologies through detailed analysis of a radiomic signature. METHODS: A radiomic model (MW2018) was fitted and externally validated using features extracted from previously reported lung and head and neck (H&N) cancer datasets using gross-tumour-volume contours, as well as from images with randomly permuted voxel index values; i.e. images without meaningful texture. To determine MW2018's added benefit, the prognostic accuracy of tumour volume alone was calculated as a baseline. RESULTS: MW2018 had an external validation concordance index (c-index) of 0.64. However, a similar performance was achieved using features extracted from images with randomized signal intensities (c-index = 0.64 and 0.60 for H&N and lung, respectively). Tumour volume had a c-index = 0.64 and correlated strongly with three of the four model features. It was determined that the signature was a surrogate for tumour volume and that intensity and texture values were not pertinent for prognostication. CONCLUSION: Our experiments reveal vulnerabilities in radiomic signature development processes and suggest safeguards that can be used to refine methodologies, and ensure productive radiomic development using objective and independent features.
PURPOSE: Refinement of radiomic results and methodologies is required to ensure progression of the field. In this work, we establish a set of safeguards designed to improve and support current radiomic methodologies through detailed analysis of a radiomic signature. METHODS: A radiomic model (MW2018) was fitted and externally validated using features extracted from previously reported lung and head and neck (H&N) cancer datasets using gross-tumour-volume contours, as well as from images with randomly permuted voxel index values; i.e. images without meaningful texture. To determine MW2018's added benefit, the prognostic accuracy of tumour volume alone was calculated as a baseline. RESULTS: MW2018 had an external validation concordance index (c-index) of 0.64. However, a similar performance was achieved using features extracted from images with randomized signal intensities (c-index = 0.64 and 0.60 for H&N and lung, respectively). Tumour volume had a c-index = 0.64 and correlated strongly with three of the four model features. It was determined that the signature was a surrogate for tumour volume and that intensity and texture values were not pertinent for prognostication. CONCLUSION: Our experiments reveal vulnerabilities in radiomic signature development processes and suggest safeguards that can be used to refine methodologies, and ensure productive radiomic development using objective and independent features.
Authors: Anshu K Jain; Sanjay Aneja; Clifton D Fuller; Adam P Dicker; Caroline Chung; Erika Kim; Justin S Kirby; Harry Quon; Clara J K Lam; William C Louv; Chris Ahern; Ying Xiao; Todd R McNutt; Nadine Housri; Ronald D Ennis; John Kang; Ying Tang; Howard Higley; Michelle A Berny-Lang; Kevin A Camphausen Journal: JCO Clin Cancer Inform Date: 2020-08
Authors: Fabien Maldonado; Cyril Varghese; Srinivasan Rajagopalan; Fenghai Duan; Aneri B Balar; Dhairya A Lakhani; Sanja L Antic; Pierre P Massion; Tucker F Johnson; Ronald A Karwoski; Richard A Robb; Brian J Bartholmai; Tobias Peikert Journal: Eur Respir J Date: 2021-04-01 Impact factor: 16.671
Authors: Kellen Mulford; Chuyu Chen; Kathryn Dusenbery; Jianling Yuan; Matthew A Hunt; Clark C Chen; Paul Sperduto; Yoichi Watanabe; Christopher Wilke Journal: Clin Transl Radiat Oncol Date: 2021-05-08
Authors: Lin Lu; Firas S Ahmed; Oguz Akin; Lyndon Luk; Xiaotao Guo; Hao Yang; Jin Yoon; A Aari Hakimi; Lawrence H Schwartz; Binsheng Zhao Journal: Front Oncol Date: 2021-05-27 Impact factor: 6.244