Ongoing BOTLab research

SCOS Cuffless Blood-Pressure Research

Timothy leads an IRB-supervised experiment on positional signal-to-noise ratio for cuffless blood-pressure estimation, applies XGBoost to 80+ engineered cardiac-waveform features, and develops wearable benchmark-test hardware.

Multi-panel scientific figure showing SCOS optical measurement principles, cardiac waveforms, and simultaneous measurement configurations.
Institution
Boston University Biomedical Optical Technologies Lab
Advisor
Ariane Garrett
Status
Ongoing research
Date
February 2026–Present
Visit BOTLab researchView the project case study

Advisor-authored publication informing the work

Ariane Garrett, Byungchan Kim, Nil Z. Gurel, Edbert J. Sie, Benjamin K. Wilson, Francesco Marsili, John P. Forman, Naomi M. Hamburg, David A. Boas, and Darren Roblyer, “Speckle contrast optical spectroscopy for cuffless blood pressure estimation based on microvascular blood flow and volume oscillations,” Biomedical Optics Express 16, 3004–3016 (2025).

View publication (opens in a new tab)

Research question

This ongoing work examines how speckle contrast optical spectroscopy (SCOS) and cardiac-waveform analysis can support cuffless blood-pressure estimation. Timothy's current experiment investigates how measurement position affects signal-to-noise ratio, connecting experimental protocols, optical measurements, data analysis, machine learning, and wearable test hardware.

Timothy's contribution

Timothy leads an IRB-supervised experiment to optimize positional signal-to-noise ratio. His responsibilities include creating and implementing the measurement protocol, recruiting participants, taking measurements, and analyzing the data.

He applies XGBoost to predict blood pressure and classify hypertension using 80+ engineered cardiac-waveform features. He also analyzes and presents features from hypertensive and non-hypertensive subjects to identify patterns relevant to cuffless blood-pressure estimation.

Using Autodesk Inventor and additive manufacturing, Timothy develops an attachable base for wireless wearable benchmark testing.

Advisor-supported workflow

Earlier analysis used a project-specific dataset that Timothy constructed from raw scientific-trial data produced through studies run by postdoctoral advisor Ariane Garrett. That advisor-run data collection is distinct from the positional experiment Timothy now leads.

Under Garrett's advisement and with advisor-assisted code, Timothy adapted an existing XGBoost workflow for his dataset. His contribution includes experimental protocol development, dataset preparation, feature analysis, cohort comparison, and project-specific adaptation.

Positional measurements and waveform methods

The current experiment pairs a defined measurement protocol with participant recruitment, measurements, and data analysis under IRB supervision. Its aim is to optimize positional signal-to-noise ratio for cuffless blood-pressure estimation.

The analytical work uses 80+ engineered cardiac-waveform features and compares hypertensive and non-hypertensive subjects to identify relevant feature patterns. Participant-level data remains private.

Machine-learning workflow

Timothy applies XGBoost to blood-pressure prediction and hypertension classification within an existing advisor-supported workflow. The work combines project-specific data preparation and feature analysis with the established modeling pipeline.

Wearable test-base development

Autodesk Inventor and additive manufacturing support an attachable base for wireless wearable benchmark testing. This work connects the analysis to practical measurement hardware.

Current status

The research and positional experiment are ongoing. Preliminary findings have been presented internally. No quantitative model-performance result or clinical-validation claim is presented here.