Machine Learning and Statistics Research Assistant
Nano-medicine Center (NEU)

During my tenure as a research assistant at the lab, I contributed to a project focused on leveraging advanced methodologies in neuroimaging analysis. We developed an automated methodology that utilized support vector machine (SVM) classification on whole-brain anatomical magnetic resonance imaging data. The primary aim was to distinguish between individuals diagnosed with Alzheimer’s disease and healthy elderly subjects. Our approach achieved an impressive mean correct classification rate of 83%, demonstrating its efficacy in effectively discerning between the two groups. Additionally, we engineered a segmentation model capable of autonomously partitioning 3D MR images into regions of interest (ROIs). This model successfully extracted crucial gray matter characteristics, significantly augmenting our understanding and analytical capabilities concerning the data. These advancements represent substantial progress in harnessing cutting-edge technology to facilitate more accurate diagnoses and foster comprehensive insights into neurological conditions.