Sebastian R van der Voort1, Fatih Incekara2,3, Maarten M J Wijnenga4, Georgios Kapsas2, Renske Gahrmann2, Joost W Schouten3, Rishi Nandoe Tewarie5, Geert J Lycklama6, Philip C De Witt Hamer7, Roelant S Eijgelaar7, Pim J French4, Hendrikus J Dubbink8, Arnaud J P E Vincent3, Wiro J Niessen1,9, Martin J van den Bent4, Marion Smits2, Stefan Klein1. 1. Biomedical Imaging Group Rotterdam, Department of Radiology and Nuclear Medicine, Erasmus MC University Medical Centre Rotterdam, Rotterdam, the Netherlands. 2. Department of Radiology and Nuclear Medicine, Erasmus MC University Medical Centre Rotterdam, Rotterdam, the Netherlands. 3. Department of Neurosurgery, Brain Tumor Center, Erasmus MC University Medical Centre Rotterdam, Rotterdam, the Netherlands. 4. Department of Neurology, Brain Tumor Center, Erasmus MC Cancer Institute, Rotterdam, the Netherlands. 5. Department of Neurosurgery, Haaglanden Medical Center, the Hague, the Netherlands. 6. Department of Radiology, Haaglanden Medical Center, the Hague, the Netherlands. 7. Department of Neurosurgery, Cancer Center Amsterdam, Brain Tumor Center, Amsterdam UMC, Amsterdam, Netherlands. 8. Department of Pathology, Brain Tumor Center at Erasmus MC Cancer Institute, Rotterdam, the Netherlands. 9. Imaging Physics, Faculty of Applied Sciences, Delft University of Technology, Delft, the Netherlands.
Abstract
BACKGROUND: Accurate characterization of glioma is crucial for clinical decision making. A delineation of the tumor is also desirable in the initial decision stages but is time-consuming. Previously, deep learning methods have been developed that can either non-invasively predict the genetic or histological features of glioma, or that can automatically delineate the tumor, but not both tasks at the same time. Here, we present our method that can predict the molecular subtype and grade, while simultaneously providing a delineation of the tumor. METHODS: We developed a single multi-task convolutional neural network that uses the full 3D, structural, pre-operative MRI scans to predict the IDH mutation status, the 1p/19q co-deletion status, and the grade of a tumor, while simultaneously segmenting the tumor. We trained our method using a patient cohort containing 1508 glioma patients from 16 institutes. We tested our method on an independent dataset of 240 patients from 13 different institutes. RESULTS: In the independent test set we achieved an IDH-AUC of 0.90, an 1p/19q co-deletion AUC of 0.85, and a grade AUC of 0.81 (grade II/III/IV). For the tumor delineation, we achieved a mean whole tumor DICE score of 0.84. CONCLUSIONS: We developed a method that non-invasively predicts multiple, clinically relevant features of glioma. Evaluation in an independent dataset shows that the method achieves a high performance and that it generalizes well to the broader clinical population. This first of its kind method opens the door to more generalizable, instead of hyper-specialized, AI methods.
BACKGROUND: Accurate characterization of glioma is crucial for clinical decision making. A delineation of the tumor is also desirable in the initial decision stages but is time-consuming. Previously, deep learning methods have been developed that can either non-invasively predict the genetic or histological features of glioma, or that can automatically delineate the tumor, but not both tasks at the same time. Here, we present our method that can predict the molecular subtype and grade, while simultaneously providing a delineation of the tumor. METHODS: We developed a single multi-task convolutional neural network that uses the full 3D, structural, pre-operative MRI scans to predict the IDH mutation status, the 1p/19q co-deletion status, and the grade of a tumor, while simultaneously segmenting the tumor. We trained our method using a patient cohort containing 1508 glioma patients from 16 institutes. We tested our method on an independent dataset of 240 patients from 13 different institutes. RESULTS: In the independent test set we achieved an IDH-AUC of 0.90, an 1p/19q co-deletion AUC of 0.85, and a grade AUC of 0.81 (grade II/III/IV). For the tumor delineation, we achieved a mean whole tumor DICE score of 0.84. CONCLUSIONS: We developed a method that non-invasively predicts multiple, clinically relevant features of glioma. Evaluation in an independent dataset shows that the method achieves a high performance and that it generalizes well to the broader clinical population. This first of its kind method opens the door to more generalizable, instead of hyper-specialized, AI methods.