Pearls in vision health research with special data science implications, as part of ADSA 2023 "Learning from data in complex and heterogeneous biological systems" session
This presentation seeks to inform data scientists of special opportunities and challenges in using electronic health record (EHR) and ophthalmic fundus image data in conducting vision health research. The author leverages specific examples from a variety of his research areas, including: public health research on vision loss in Southwest Virginia; generalizable video AI/ML models to democratize access to glaucoma screening; and investigating health disparities in glaucoma progression using multimodal, explainable artificial intelligence.
Learning objectives of this talk include: potential vision health applications of data science methods; nature and limitations of vision related epidemiologic data; how “Big Data” from multiple sources can inform the assessment of visual impairment at a population level; health disparities and special vision related considerations in rural and underserved regions; the contribution of quantitative and qualitative data in triangulating visual impairment and access to care in developing vision health programs; how data science and large bioinformatics databases can better inform vision related investigations, including vision public health .