Dr. Linn's methodological research addresses barriers to incorporating complex, high-dimensional data into models for personalized medicine and biomarker development using neuroimaging data. As a co-investigator, she provides biostatistical leadership on projects in behavioral health economics, implementation science, and psychiatry/psychology. 

Dr. Linn has experience designing sequential multiple assignment randomized trials (SMARTs) in a number of research areas. She is particularly interested in using data from SMART designs to estimate personalized dynamic interventions that improve long-term outcomes for individuals. Her methodological research in this area aims to address practical issues such as how to handle missing data when estimating optimal individualized treatment rules and evaluating their expected performance in a target population.

Dr. Linn has made contributions to the field of neuroimaging statistics through methods for harmonizing imaging data from multi-site studies, mitigating the influence of nuisance variables in image-based biomarker development, and estimating spatially varying patterns of multimodal image associations. She is interested in adapting methods from causal inference to address confounding in analyses of imaging data and personalizing the treatment of depression using fMRI-guided Transcranial Magnetic Stimulation (TMS) protocols.