H Zhou1, R Hu2, O Tang3, C Hu4, L Tang1, K Chang5, Q Shen6, J Wu6, B Zou2, B Xiao1, J Boxerman7, W Chen8, R Y Huang9, L Yang10, H X Bai7, C Zhu11,12,13. 1. Department of Neurology (H.Z., L.T., B.X.), Xiangya Hospital of Central South University, Changsha, Hunan, China. 2. From the School of Computer Science and Engineering (R.H., B.Z., C.Z.). 3. Warren Alpert Medical School, Brown University (O.T.), Providence, Rhode Island. 4. Department of Neurology (C.H.), Hunan Provincial People's Hospital, Changsha, Hunan, China. 5. Department of Radiology (K.C.), Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, Harvard Medical School, Boston, Massachusetts. 6. Radiology (Q.S., J.W.), Second Xiangya Hospital of Central South University, Changsha, Hunan, China. 7. Department of Diagnostic Imaging (J.B., H.X.B.), Rhode Island Hospital. 8. Department of Pathology (W.C.), Hunan Children's Hospital, Changsha, Hunan, China. 9. Department of Radiology (R.Y.H.), Brigham and Women's Hospital, Boston, Massachusetts. 10. Departments of Neurology (L.Y.). 11. From the School of Computer Science and Engineering (R.H., B.Z., C.Z.) anandawork@126.com. 12. College of Literature and Journalism (C.Z.), Central South University, Changsha, Hunan, China. 13. Mobile Health Ministry of Education-China Mobile Joint Laboratory (C.Z.), China.
Abstract
BACKGROUND AND PURPOSE: Differentiating the types of pediatric posterior fossa tumors on routine imaging may help in preoperative evaluation and guide surgical resection planning. However, qualitative radiologic MR imaging review has limited performance. This study aimed to compare different machine learning approaches to classify pediatric posterior fossa tumors on routine MR imaging. MATERIALS AND METHODS: This retrospective study included preoperative MR imaging of 288 patients with pediatric posterior fossa tumors, including medulloblastoma (n = 111), ependymoma (n = 70), and pilocytic astrocytoma (n = 107). Radiomics features were extracted from T2-weighted images, contrast-enhanced T1-weighted images, and ADC maps. Models generated by standard manual optimization by a machine learning expert were compared with automatic machine learning via the Tree-Based Pipeline Optimization Tool for performance evaluation. RESULTS: For 3-way classification, the radiomics model by automatic machine learning with the Tree-Based Pipeline Optimization Tool achieved a test micro-averaged area under the curve of 0.91 with an accuracy of 0.83, while the most optimized model based on the feature-selection method χ2 score and the Generalized Linear Model classifier achieved a test micro-averaged area under the curve of 0.92 with an accuracy of 0.74. Tree-Based Pipeline Optimization Tool models achieved significantly higher accuracy than average qualitative expert MR imaging review (0.83 versus 0.54, P < .001). For binary classification, Tree-Based Pipeline Optimization Tool models achieved an area under the curve of 0.94 with an accuracy of 0.85 for medulloblastoma versus nonmedulloblastoma, an area under the curve of 0.84 with an accuracy of 0.80 for ependymoma versus nonependymoma, and an area under the curve of 0.94 with an accuracy of 0.88 for pilocytic astrocytoma versus non-pilocytic astrocytoma. CONCLUSIONS: Automatic machine learning based on routine MR imaging classified pediatric posterior fossa tumors with high accuracy compared with manual expert pipeline optimization and qualitative expert MR imaging review.
BACKGROUND AND PURPOSE: Differentiating the types of pediatric posterior fossa tumors on routine imaging may help in preoperative evaluation and guide surgical resection planning. However, qualitative radiologic MR imaging review has limited performance. This study aimed to compare different machine learning approaches to classify pediatric posterior fossa tumors on routine MR imaging. MATERIALS AND METHODS: This retrospective study included preoperative MR imaging of 288 patients with pediatric posterior fossa tumors, including medulloblastoma (n = 111), ependymoma (n = 70), and pilocytic astrocytoma (n = 107). Radiomics features were extracted from T2-weighted images, contrast-enhanced T1-weighted images, and ADC maps. Models generated by standard manual optimization by a machine learning expert were compared with automatic machine learning via the Tree-Based Pipeline Optimization Tool for performance evaluation. RESULTS: For 3-way classification, the radiomics model by automatic machine learning with the Tree-Based Pipeline Optimization Tool achieved a test micro-averaged area under the curve of 0.91 with an accuracy of 0.83, while the most optimized model based on the feature-selection method χ2 score and the Generalized Linear Model classifier achieved a test micro-averaged area under the curve of 0.92 with an accuracy of 0.74. Tree-Based Pipeline Optimization Tool models achieved significantly higher accuracy than average qualitative expert MR imaging review (0.83 versus 0.54, P < .001). For binary classification, Tree-Based Pipeline Optimization Tool models achieved an area under the curve of 0.94 with an accuracy of 0.85 for medulloblastoma versus nonmedulloblastoma, an area under the curve of 0.84 with an accuracy of 0.80 for ependymoma versus nonependymoma, and an area under the curve of 0.94 with an accuracy of 0.88 for pilocytic astrocytoma versus non-pilocytic astrocytoma. CONCLUSIONS: Automatic machine learning based on routine MR imaging classified pediatric posterior fossa tumors with high accuracy compared with manual expert pipeline optimization and qualitative expert MR imaging review.
Authors: D Rodriguez Gutierrez; A Awwad; L Meijer; M Manita; T Jaspan; R A Dineen; R G Grundy; D P Auer Journal: AJNR Am J Neuroradiol Date: 2013-12-05 Impact factor: 3.825
Authors: Karen A Manias; Simrandip K Gill; Lesley MacPherson; Adam Oates; Benjamin Pinkey; Paul Davies; Niloufar Zarinabad; Nigel P Davies; Ben Babourina-Brooks; Martin Wilson; Andrew C Peet Journal: Neurooncol Pract Date: 2019-05-09
Authors: Siddhi Ramesh; Sukarn Chokkara; Timothy Shen; Ajay Major; Samuel L Volchenboum; Anoop Mayampurath; Mark A Applebaum Journal: JCO Clin Cancer Inform Date: 2021-12
Authors: M Zhang; L Tam; J Wright; M Mohammadzadeh; M Han; E Chen; M Wagner; J Nemalka; H Lai; A Eghbal; C Y Ho; R M Lober; S H Cheshier; N A Vitanza; G A Grant; L M Prolo; K W Yeom; A Jaju Journal: AJNR Am J Neuroradiol Date: 2022-03-31 Impact factor: 3.825
Authors: Michael Zhang; Samuel W Wong; Jason N Wright; Sebastian Toescu; Maryam Mohammadzadeh; Michelle Han; Seth Lummus; Matthias W Wagner; Derek Yecies; Hollie Lai; Azam Eghbal; Alireza Radmanesh; Jordan Nemelka; Stephen Harward; Michael Malinzak; Suzanne Laughlin; Sebastien Perreault; Kristina R M Braun; Arastoo Vossough; Tina Poussaint; Robert Goetti; Birgit Ertl-Wagner; Chang Y Ho; Ozgur Oztekin; Vijay Ramaswamy; Kshitij Mankad; Nicholas A Vitanza; Samuel H Cheshier; Mourad Said; Kristian Aquilina; Eric Thompson; Alok Jaju; Gerald A Grant; Robert M Lober; Kristen W Yeom Journal: Neurosurgery Date: 2021-10-13 Impact factor: 5.315
Authors: Nalini M Singh; Jordan B Harrod; Sandya Subramanian; Mitchell Robinson; Ken Chang; Suheyla Cetin-Karayumak; Adrian Vasile Dalca; Simon Eickhoff; Michael Fox; Loraine Franke; Polina Golland; Daniel Haehn; Juan Eugenio Iglesias; Lauren J O'Donnell; Yangming Ou; Yogesh Rathi; Shan H Siddiqi; Haoqi Sun; M Brandon Westover; Susan Whitfield-Gabrieli; Randy L Gollub Journal: Neuroinformatics Date: 2022-03-28
Authors: Michael Zhang; Edward Wang; Derek Yecies; Lydia T Tam; Michelle Han; Sebastian Toescu; Jason N Wright; Emre Altinmakas; Eric Chen; Alireza Radmanesh; Jordan Nemelka; Ozgur Oztekin; Matthias W Wagner; Robert M Lober; Birgit Ertl-Wagner; Chang Y Ho; Kshitij Mankad; Nicholas A Vitanza; Samuel H Cheshier; Tom S Jacques; Paul G Fisher; Kristian Aquilina; Mourad Said; Alok Jaju; Stefan Pfister; Michael D Taylor; Gerald A Grant; Sarah Mattonen; Vijay Ramaswamy; Kristen W Yeom Journal: Neuro Oncol Date: 2022-06-01 Impact factor: 13.029
Authors: M Zhang; S W Wong; S Lummus; M Han; A Radmanesh; S S Ahmadian; L M Prolo; H Lai; A Eghbal; O Oztekin; S H Cheshier; P G Fisher; C Y Ho; H Vogel; N A Vitanza; R M Lober; G A Grant; A Jaju; K W Yeom Journal: AJNR Am J Neuroradiol Date: 2021-07-15 Impact factor: 4.966