Approved research

Surface based neuroimaging analysis to advance understanding of neurologic conditions

Lay summary

We propose the application of a novel approach to brain imaging analysis, surface-based deep learning (SBDL), to the study of a variety of neurologic disorders including epilepsy, Alzheimers Disease (AD), autism and chronic pain. This is an entirely novel approach to brain imaging analysis and was developed by my team at Northwestern University. Deep learning is a machine learning technique becoming increasingly prevalent due to its outstanding performance in fields such as computer vision, medical imaging and artificial intelligence. The strength of this method is to learn the most relevant features without any prior knowledge of the data. Using our method, we have been able to successfully predict both age and gender accurately in healthy control subjects. The next phase of developing this method is to extend its application into disease states in order to explore its utility in predicting clinically relevant variables and informing clinical outcomes. The expected value of this research is to advance our understanding of a variety of neurologic conditions using this novel technique for neuroimaging analysis. Given that our method has been used to successfully predict demographic information in healthy controls, we therefore believe it will be robust in its ability to predict clinically relevant disease variables and clinical outcomes in patients suffering from neurologic conditions. We propose an initial 3 year period of time to explore the use of our method to a variety of neurologic conditions. On a longer term basis, these investigations have the potential to significantly impact our understanding of these conditions, inform patient and family counseling at the time of diagnosis and identify novel treatment targets in these conditions which collectively affect hundreds of millions of individuals around the world.