Research Library

PUBLISHED: OCT 2025 • ALGORITHM Deep Learning Approaches to Endothelial Function Prediction

We propose a novel Convolutional Neural Network (CNN) architecture designed to extract vascular reactivity features from noisy photoplethysmography (PPG) signals.

DRAFT • MACHINE LEARNING Real-time Anomaly Detection on Edge Devices (nRF52840)

Optimizing TinyML models to run directly on the Nordic nRF52840 SoC, enabling 24/7 arrhythmia detection without cloud dependency.

WHITE PAPER • ENGINEERING Active Noise Cancellation in Ambulatory ECG

A breakdown of our proprietary circuit design that subtracts motion artifacts from the ECG signal path using dual-reference sensors.

CLINICAL TRIAL • PHASE 1 Comparative Analysis: Aoknos vs. Brachial Artery Ultrasound

Results from our 500-patient study comparing AI-predicted endothelial function against the gold-standard ultrasound method.