Topological Data Analysis on Magnetic Resonance Image Biomarkers

Big data provides us the source to extract new insights in various disciplines. In healthcare, big data may help discover the pathology of a disease and new effective treatments. However, the era of big data taxes the ability of many researchers to analyze and interact with data in biomedical research. Traditional statistical methods have limitations to infer relevant features for possibly complex data. Topological Data Analysis (TDA) provides a set of new topological and geometric tools to discover the hidden relations or key features from complex data. It may help identify the key risk factors related to a given disease and reduce the noise impact from other factors. In this paper, we use TDA to analyze the multidimensional data from MRI of knee osteoarthritis patients.