Raindrop animation: Visualizing change in longitudinal biomarker data. (20th December 2022)

• Record Type: Journal Article
• Authors: Saef, Benjamin A.; Henson, Rachel L.; Volluz, Katherine; Yarasheski, Kevin E.; West, Tim; Kirmess, Kris; Meyer, Matthew R.; Gordon, Brian A.; Benzinger, Tammie L.S.; Morris, John C.; Fagan, Anne M.; Schindler, Suzanne E.
• Is Part Of: Alzheimer's & dementia. Volume 18(2022)Supplement 5
• Page Start: n/a
• Page End: n/a
• Publication Date: 2022-12-20
• Subjects: Alzheimer's disease -- Periodicals; Alzheimer Disease -- Periodicals; Dementia -- Periodicals; Démence; Maladie d'Alzheimer; Périodique électronique (Descripteur de forme); Ressource Internet (Descripteur de forme); 616.83
• DOI: 10.1002/alz.067300
• Languages: English
• ISSNs: 1552-5260

Abstract

Abstract: Background: The Knight ADRC has collected fluid biomarker data from some participants over more than >15 years, affording a unique opportunity to explore trends in the progression of Alzheimer disease biomarkers. Evaluation of intra‐individual change provides a uniquely powerful approach to understanding the development of Alzheimer disease pathophysiology, notably non‐linear changes in biomarkers. To facilitate understanding of longitudinal analyses, an approach was developed to visualize longitudinal change in biomarkers. Method: Samples from ADRC participants enrolled in studies of memory and aging were analyzed for CSF Aβ40, Aβ42, total tau, and phospho‐tau181 using the fully automated LUMIPULSE assay platform (Fujirebio). Plasma Aβ40 and Aβ42 were measured via a high‐precision immunoprecipitation‐mass spectrometry assay (C2N Diagnostics). Brain amyloid burden was measured within two years of fluid collection via amyloid PET. Biomarker values between measurements were estimated by linear regression. A Shiny application was written in R to visualize estimated longitudinal biomarker values on a granular time scale. Result: The R Shiny and ggplot2 packages were used to create interactive animated plots of biomarker changes over time. This application utilizes CSF, plasma, and amyloid PET data to display within‐participant longitudinal change of the selected biomarkers as a function of other baseline biomarker measurements. The estimated biomarker value at a timepoint after the baseline is shown as a larger point, and the estimated values from the previous year are shown as smaller points, giving the visual appearance of one raindrop for each individual. The interface allows for users to select the biomarkers represented on each axis, color coding by longitudinal or baseline biomarker values, and biomarker cut‐offs. The user can animate the customized graph to show changes over >15 years of accumulated longitudinal data. Conclusion: Our web‐based application allows for visualization of biomarker changes over time. This tool allows for comparison of CSF, plasma, and imaging markers in an interactive format, facilitating the creation of novel hypotheses by observation of longitudinal trends.