An Exploratory Study of Blood Glucose Estimation from Photoplethysmography Signals using Machine Learning
Researchers have explored the use of machine learning and deep learning to estimate blood glucose levels non-invasively using photoplethysmogram (PPG) signals from smartwatches. This approach aims to overcome the limitations of traditional, invasive continuous glucose monitoring (CGM) devices, which can cause irritation. The study presents a paired dataset of PPG signals and CGM data, with preliminary results indicating potential predictive signals, though further research with larger datasets is necessary. AI
IMPACT This research could lead to non-invasive blood glucose monitoring, improving diabetes management through wearable technology.