51. Predicting pediatric optic pathway glioma progression using advanced magnetic resonance image analysis and machine learning
- Author
-
Martin Rozycki, Jayesh P. Thawani, Gregory G. Heuer, Arastoo Vossough, Christos Davatzikos, Phillip B. Storm, Michael Fisher, Robert A. Avery, Jared M Pisapia, and Hamed Akbari
- Subjects
Radiography ,diffusion-weighted imaging ,Clinical Investigations ,Machine learning ,computer.software_genre ,White matter ,03 medical and health sciences ,0302 clinical medicine ,Fractional anisotropy ,Medical imaging ,medicine ,AcademicSubjects/MED00300 ,magnetic resonance imaging ,Diffusion Tractography ,medicine.diagnostic_test ,optic pathway glioma ,business.industry ,Magnetic resonance imaging ,medicine.anatomical_structure ,machine learning ,Tumor progression ,030220 oncology & carcinogenesis ,AcademicSubjects/MED00310 ,Artificial intelligence ,business ,computer ,030217 neurology & neurosurgery ,vision decline ,Diffusion MRI - Abstract
Background Optic pathway gliomas (OPGs) are low-grade tumors of the white matter of the visual system with a highly variable clinical course. The aim of the study was to generate a magnetic resonance imaging (MRI)-based predictive model of OPG tumor progression using advanced image analysis and machine learning techniques. Methods We performed a retrospective case–control study of OPG patients managed between 2009 and 2015 at an academic children’s hospital. Progression was defined as radiographic tumor growth or vision decline. To generate the model, optic nerves were manually highlighted and optic radiations (ORs) were segmented using diffusion tractography tools. For each patient, intensity distributions were obtained from within the segmented regions on all imaging sequences, including derivatives of diffusion tensor imaging (DTI). A machine learning algorithm determined the combination of features most predictive of progression. Results Nineteen OPG patients with progression were matched to 19 OPG patients without progression. The mean time between most recent follow-up and most recently analyzed MRI was 3.5 ± 1.7 years. Eighty-three MRI studies and 532 extracted features were included. The predictive model achieved an accuracy of 86%, sensitivity of 89%, and specificity of 81%. Fractional anisotropy of the ORs was among the most predictive features (area under the curve 0.83, P < 0.05). Conclusions Our findings show that image analysis and machine learning can be applied to OPGs to generate a MRI-based predictive model with high accuracy. As OPGs grow along the visual pathway, the most predictive features relate to white matter changes as detected by DTI, especially within ORs.
- Published
- 2020