Machine learning based generation of ontology for structural and functional mapping

This invention uses machine learning to create a structured map, called an ontology, that links specific brain areas to various mental functions. It works by analyzing large amounts of text describing mental functions and spatial data about brain structures. One machine learning step identifies initial connections, and another refines the map by optimizing the number of categories and terms within them. This map can then be used to analyze a patient's electronic medical records and predict their health outcomes. The claims specify that the first machine learning step can be an unsupervised technique like k-means clustering, looking for co-occurrences between brain structures and mental function terms.