1
|
Iv M, Zhou M, Shpanskaya K, Perreault S, Wang Z, Tranvinh E, Lanzman B, Vajapeyam S, Vitanza NA, Fisher PG, Cho YJ, Laughlin S, Ramaswamy V, Taylor MD, Cheshier SH, Grant GA, Young Poussaint T, Gevaert O, Yeom KW. MR Imaging-Based Radiomic Signatures of Distinct Molecular Subgroups of Medulloblastoma. AJNR Am J Neuroradiol 2018; 40:154-161. [PMID: 30523141 DOI: 10.3174/ajnr.a5899] [Citation(s) in RCA: 66] [Impact Index Per Article: 11.0] [Reference Citation Analysis] [What about the content of this article? (0)] [Affiliation(s)] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 06/29/2018] [Accepted: 10/06/2018] [Indexed: 12/12/2022]
Abstract
BACKGROUND AND PURPOSE Distinct molecular subgroups of pediatric medulloblastoma confer important differences in prognosis and therapy. Currently, tissue sampling is the only method to obtain information for classification. Our goal was to develop and validate radiomic and machine learning approaches for predicting molecular subgroups of pediatric medulloblastoma. MATERIALS AND METHODS In this multi-institutional retrospective study, we evaluated MR imaging datasets of 109 pediatric patients with medulloblastoma from 3 children's hospitals from January 2001 to January 2014. A computational framework was developed to extract MR imaging-based radiomic features from tumor segmentations, and we tested 2 predictive models: a double 10-fold cross-validation using a combined dataset consisting of all 3 patient cohorts and a 3-dataset cross-validation, in which training was performed on 2 cohorts and testing was performed on the third independent cohort. We used the Wilcoxon rank sum test for feature selection with assessment of area under the receiver operating characteristic curve to evaluate model performance. RESULTS Of 590 MR imaging-derived radiomic features, including intensity-based histograms, tumor edge-sharpness, Gabor features, and local area integral invariant features, extracted from imaging-derived tumor segmentations, tumor edge-sharpness was most useful for predicting sonic hedgehog and group 4 tumors. Receiver operating characteristic analysis revealed superior performance of the double 10-fold cross-validation model for predicting sonic hedgehog, group 3, and group 4 tumors when using combined T1- and T2-weighted images (area under the curve = 0.79, 0.70, and 0.83, respectively). With the independent 3-dataset cross-validation strategy, select radiomic features were predictive of sonic hedgehog (area under the curve = 0.70-0.73) and group 4 (area under the curve = 0.76-0.80) medulloblastoma. CONCLUSIONS This study provides proof-of-concept results for the application of radiomic and machine learning approaches to a multi-institutional dataset for the prediction of medulloblastoma subgroups.
Collapse
Affiliation(s)
- M Iv
- From the Department of Radiology (M.I., M.Z., K.S., E.T., B.L., K.W.Y.)
| | - M Zhou
- From the Department of Radiology (M.I., M.Z., K.S., E.T., B.L., K.W.Y.).,Stanford Center for Biomedical Informatics (M.Z., O.G., Z.W.)
| | - K Shpanskaya
- From the Department of Radiology (M.I., M.Z., K.S., E.T., B.L., K.W.Y.)
| | - S Perreault
- Department of Pediatrics (S.P.), Pediatric Neurology, Centre Hospitalier Universitaire Sainte Justine, University of Montréal, Montreal, Quebec, Canada
| | - Z Wang
- Stanford Center for Biomedical Informatics (M.Z., O.G., Z.W.)
| | - E Tranvinh
- From the Department of Radiology (M.I., M.Z., K.S., E.T., B.L., K.W.Y.)
| | - B Lanzman
- From the Department of Radiology (M.I., M.Z., K.S., E.T., B.L., K.W.Y.)
| | - S Vajapeyam
- Department of Radiology (S.V., T.Y.P.), Boston Children's Hospital, Harvard University, Boston, Massachusetts
| | - N A Vitanza
- Department Pediatrics Hematology-Oncology (N.A.V.), Seattle Children's Hospital, University of Washington, Seattle, Washington
| | - P G Fisher
- Department of Pediatrics (P.G.F.), Pediatric Neurology
| | - Y J Cho
- Department of Pediatrics (Y.J.C.), Pediatric Neurology, Oregon Health & Science University, Portland, Oregon
| | - S Laughlin
- Departments of Radiology, Neuro-Oncology, and Neurosurgery (S.L., V.R., M.D.T.), Hospital for Sick Children, Toronto, Ontario, Canada
| | - V Ramaswamy
- Departments of Radiology, Neuro-Oncology, and Neurosurgery (S.L., V.R., M.D.T.), Hospital for Sick Children, Toronto, Ontario, Canada
| | - M D Taylor
- Departments of Radiology, Neuro-Oncology, and Neurosurgery (S.L., V.R., M.D.T.), Hospital for Sick Children, Toronto, Ontario, Canada
| | - S H Cheshier
- Department of Neurosurgery (S.H.C.), Pediatric Neurosurgery, University of Utah, Salt Lake City, Utah
| | - G A Grant
- Department of Neurosurgery (G.A.G.), Pediatric Neurosurgery, Lucile Packard Children's Hospital, Stanford University, Palo Alto, California
| | - T Young Poussaint
- Department of Radiology (S.V., T.Y.P.), Boston Children's Hospital, Harvard University, Boston, Massachusetts
| | - O Gevaert
- Stanford Center for Biomedical Informatics (M.Z., O.G., Z.W.)
| | - K W Yeom
- From the Department of Radiology (M.I., M.Z., K.S., E.T., B.L., K.W.Y.) .,Department of Radiology (K.W.Y.), Artificial Intelligence in Medicine and Imaging, Stanford University, Stanford, California
| |
Collapse
|