Development and Internal Validation of an AI-Enabled Cuff-less, Non-invasive Continuous Blood Pressure Monitor Across All Classes of Hypertension

Francisco Lopez-Jimenez; Abhishek Deshmukh; John Bisognano; John Boehmer; Mouli Ramasamy; Prashanth Shyam Kumar; Suraj Kapa; Venk Varadan; Vijay Varadan; Marat Fudim

doi:10.1007/s12265-024-10589-5

Development and Internal Validation of an AI-Enabled Cuff-less, Non-invasive Continuous Blood Pressure Monitor Across All Classes of Hypertension

J Cardiovasc Transl Res. 2025 Jan 13. doi: 10.1007/s12265-024-10589-5. Online ahead of print.

Authors

Affiliations

¹ Department of Cardiovascular Medicine, Mayo Clinic, Rochester, MN, 55905, USA.
² Department of Internal Medicine, Department of Cardiology, University of Michigan, Ann Arbor, MI, 48109, USA.
³ Department of Medicine, Penn State Heart and Vascular Institute, Hershey, PA, 17033, USA.
⁴ Nanowear Inc, Brooklyn, NY, 11201, USA.
⁵ Duke University Medical Center, Durham, NC, 27710, USA. marat.fudim@gmail.com.
⁶ Duke Clinical Research Institute, Durham, NC, 27710, USA. marat.fudim@gmail.com.

PMID: 39804564
DOI: 10.1007/s12265-024-10589-5

Abstract

Background: Non-invasive, continuous blood pressure monitoring technologies require additional validation beyond standard cuff-based methods. This study evaluates a non-invasive, multiparametric wearable cuffless blood pressure (BP) diagnostic monitor across all hypertension classes with diverse subjects.

Methods: A prospective, multicenter study assessed Nanowear's SimpleSense-BP performance, including induced and natural BP changes, significant BP variations (Systolic BP (SBP) ≥ ± 15 mm Hg and Diastolic BP (DBP) ≥ ± 10 mm Hg), and reference input value validity over 4 weeks.

Results: 303 subjects (18-83 yrs; 50.16% Female) participated in algorithmic development and validation (Normal - 35%, Prehypertensive - 24%, Stage 1 - 24%, Stage 2 - 17%). 54 subjects were tested for induced change performance, 149 exhibited significant changes, and 91 validated reference value duration.

Conclusions: The study clinically validated a continuous, AI-based BP diagnostic monitor using non-invasive wearable data. Further testing on diverse populations and external validation are recommended. The protocol was inspired by ISO 81060-2 and IEEE 1708:2019 standards.

Keywords: Ambulatory; Artificial intelligence; Blood pressure; Diagnostic; Hypertension.