Barrow Neurological Institute
Phoenix, Arizona, United States
I am currently serving as an Associated Professor in the Department of Translational Neuroscience and a director of the Biostatistics Program, Ivy Brain Tumor Center at Barrow Neurological Institute. As a cancer biostatistician as well as a population health scientist over last twenty years, I have developed statistical methodologies and applications on pre-clinical and clinical trials (early phases) and observational studies as translational and clinical research with inter- and multi-disciplinary research teams. In the proposed study, we compared the predictive performance of various bagging models using artificial neural network (ANN) methods on GBM patients within summary data and explore the benefits of ensemble learning for reducing overfitting and improving generalizability from limited summary data with the goal of creating a methodology for researchers to assess single-arm experimental efficacy.
My current research includes study designs of phase 0 trial in glioblastoma (GBM), methodological development in phase 2 single-arm clinical trials in high-risk oncology trials, and data-driven approach using machine-learning algorithm with large cancer and neuroscience health database. Here are selected research papers related to my research interests: (1) My recent paper (https://pubmed.ncbi.nlm.nih.gov/36550391/) evaluated the two-stage designs in GBM through the systematic review and discussed the design issues related to single-arm studies in GBM. (2) A data-driven paper (https://pubmed.ncbi.nlm.nih.gov/32345710/) utilized a longitudinal data-based approach to examine state-level human papillomavirus (HPV) vaccine trends and their influences over time. Growth mixture modeling was used to analyze the National Immunization Study (NIS:2008-2016) database. (3) A research paper (https://pubmed.ncbi.nlm.nih.gov/23795347/) compared the detection power of four methods of logistic regression, logic regression, classification trees, and random forests to identify effective gene-gene and gene-environmental interactions using a large-scale population-based case-control study. (4) A health disparity paper (https://pubmed.ncbi.nlm.nih.gov/28234949/) on cervical cancer incidence and mortality for older women. (5) A multicollinearity study (https://pubmed.ncbi.nlm.nih.gov/25664257/).
