Abstracts

Rationale: Cortical dysplasia is a congenital migrational disorder of brain development with a prevalence between 5 and 25% among patients diagnosed with epilepsy. When cortical dysplasia causes epilepsy, the seizures are often refractory to antiepileptic medications and surgery is often the best treatment option for these patients. Thus, the identification of cortical dysplasia can be an important factor in a patient's treatment plan.Identification of cortical dysplasia is typically done by radiologists looking through multiple MRI scans and sequences across multiple planes. The review can be time consuming even for experienced neuroradiologists and may sometimes still not detect cortical dysplasia. Recent advances in machine learning (ML) methods have significantly improved computerized image processing for object recognition and make it possible to segment different parts of an image and correctly label each segment. That is, computers can now 'look' at a picture and tell if it's a picture of a cat, a dog, or both, for example.  Methods: For this pilot project, the preoperative axial FLAIR MRIs from 18 pediatric patients who were evaluated in the EMU at Dell Children's Medical Center in Austin, TX, were used. Patients ranged from 2 to 19 years of age. The MRIs were acquired on a 1.5T Siemens magnet during the surgical evaluation for each patient and the axial FLAIR sequence was selected as a common sequence across all patients. Slices in the axial FLAIR sequence were 5 mm thick and spacing was 6.25 mm.In order for the ML system to correctly segment and label an image, the ML model must be trained on a set of images that have been annotated to indicate to the computer model what regions of the image show dysplasia. The images for this pilot study were annotated based on the radiological report, postsurgical MRI images, and pathology reports for each patient. There were 1632 axial FLAIR images used to train the model. Each image was centered in a 256 by 256 voxel array.The ML model was built in Python 3.6 using the Tensorflow platform. The ML architecture was a U-Net convolutional neural network. The U-Net architecture involves a contraction path to extract advanced features in the image, but also reduces the size of feature maps. Thus, an expansion path is needed to recover the size of the feature map. Finally, a concatenation path is used to combine the contraction and expansion feature maps to generate a complete feature map for segmentation and labeling of the image.  Results: The model was trained over 500 epochs of the data set. The model training accuracy was 97.05% and the model validation accuracy was 98.50%. Figure 1 shows that the model training and validation accuracy converged over the course of training, as expected.  Conclusions: This pilot study, with a small patient sample and only one image sequence per patient, shows that machine learning has the potential to segment MR images and label segments showing cortical dysplasia. This, in turn, has the potential to identify cortical dysplasia in patients faster and improve their treatments. Additional discussion will cover the potential for generalizing this model across different MRI sequences and slice axes.  Funding: No funding