Dr. Shanmugathasan Suthaharan (Computer Science) received new funding from the University of Pittsburgh for the project “Research Collaboration – Graduate Student Workers at UNCG.”
The main goal of this project is to study multimodal longitudinal ophthalmic images and develop feature learning models that can be integrated into Deep Temporal Networks (DTNets) — a neuromorphic engineered machine learning framework — to detect and classify age-related macular degeneration (AMD) retinal diseases. Three students from UNC Greensboro will work with Dr. Shan Suthaharan (UNCG-PI) in the design and development of feature learning models and algorithms, and fully engaged in the implementation and testing of software modules for the algorithms using Matlab, R, or Python.
More specifically, one graduate student will study multi-modal ophthalmic images, work with Dr. Suthaharan on the development of computational models and algorithms, and implement the software modules (including data fusion techniques) that are suitable for integrating them into DTNets for detecting and characterizing retinal diseases. The second graduate student will study the OCT retinal images and work with Dr. Suthaharan for implementing the algorithms (morphological and topological) that help the extraction of latent features of retinal diseases. The third graduate student will work with Dr. Suthaharan and develop algorithms that identify and isolate overlapping features (data heterogeneity) detected in multi-modal ophthalmic images. Then implement algorithms to separate AMD disease signatures using these isolated and labeled features.