PT  - JOURNAL ARTICLE
AU  - Zhou, H.
AU  - Hu, R.
AU  - Tang, O.
AU  - Hu, C.
AU  - Tang, L.
AU  - Chang, K.
AU  - Shen, Q.
AU  - Wu, J.
AU  - Zou, B.
AU  - Xiao, B.
AU  - Boxerman, J.
AU  - Chen, W.
AU  - Huang, R.Y.
AU  - Yang, L.
AU  - Bai, H.X.
AU  - Zhu, C.
TI  - Automatic Machine Learning to Differentiate Pediatric Posterior Fossa Tumors on Routine MR Imaging
AID  - 10.3174/ajnr.A6621
DP  - 2020 Jul 01
TA  - American Journal of Neuroradiology
PG  - 1279--1285
VI  - 41
IP  - 7
4099  - http://www.ajnr.org/content/41/7/1279.short
4100  - http://www.ajnr.org/content/41/7/1279.full
SO  - Am. J. Neuroradiol.2020 Jul 01; 41
AB  - 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.AUCarea under the curveAutoMLautomatic machine learningCHSQχ2 scoreEPependymomaMBmedulloblastomaMLmachine learningPApilocytic astrocytomaTPOTTree-Based Pipeline Optimization Tool