January 23, 2025

About the speaker

Roberto Arias is a PhD Candidate at the University of Texas Rio Grande Valley, studying Mathematics & Statistics with Interdisciplinary Applications since Fall 2022. He graduated from the SACM at Texas A&M University Kingsville. He is passionate about teaching and receives wonderful reviews from his students on the basis of being well-organized, understanding, and willing to explain the mathematics step-by-step. He is an Editorial Board Member of the Journal titled Model Assisted Statistics and Applications. After graduation, Roberto looks forward to a long career in statistical analysis. 

Novel Feature Extraction from Discrete 24H Ambulatory Blood Pressure

Ambulatory blood pressure (ABP) monitoring provides 24-hour BP measurements, critical for diagnosing hypertension, assessing cardiovascular risk, and identifying abnormalities often undetected in conventional checks. Traditional ABP metrics focus on averages or linear measures, overlooking the nonlinear, multi-scale nature of BP regulation. This study uses 24-hour ABP data, recorded at 15-minute intervals during the day and 30-minute intervals at night, to extract novel features offering deeper clinical insights. By applying Maximal Overlap Discrete Wavelet Transform (MODWT) on interpolated ABP data, BP time series are decomposed into multiple resolution levels, revealing patterns across various temporal scales. Recurrence Quantification Analysis (RQA) metrics such as recurrence rate, determinism, and laminarity capture dynamic, nonlinear BP variability and behavior. Additionally, the derivative of interpolated data identifies abrupt surges or dips, emphasizing transient dynamics. These features are integrated into machine learning frameworks along with mixed effect modeling to assess their predictive value for health risks. This innovative approach provides time-localized, scale-specific insights into BP regulation, addressing limitations of traditional methods and enabling more refined clinical assessments. The methodology demonstrates significant potential for enhancing ABPM interpretation and advancing tailored treatment strategies, offering a transformative perspective on BP variability.