Non-invasive Measurement of Glucose with Advanced Spectroscopy Techniques and Machine Learning Models

A Pilot Study to Assess the Ability of Advanced Spectroscopy Techniques and Machine Learning Models to Non-invasively Measure Glucose Robustly and Accurately

Primary aim of this pilot study is to provide evidence for a non-invasive, continuous, and transcutaneous measurement method to determine blood glucose levels robustly and accurately.

The study is divided into 3 sub-studies focusing on

• Interchangeability of Lab demo 1.0
• Evaluation of advanced spectroscopy methods and their miniaturisation potential
• Robustness of machine learning model to changes over time

Inclusion Criteria:

• Provide signed and dated written informed consent by study participant prior to any mandatory study-specific procedures, sample collection, or analysis
• Apparently healthy Male or Female adult aged between 18 and 75
• Willingness to follow the study procedures

Exclusion Criteria:

• In Female subjects: pregnancy or breastfeeding period
• A diagnosis of diabetes mellitus (any type) with ongoing insulin therapy (basal or basal plus bolus) or sulfonylurea therapy
• Prior severe hypoglycemia event (stage 3 or above; e.g., seizure, loss of consciousness, emergency hospitalisation)
• Prior event of severe hyperglycemia resulting in diabetic ketoacidosis or hyperosmolar hyperglycemic state, or emergency hospitalisation
• Documented medical history of bleeding disorder
• Being under anticoagulant medication
• Insufficient knowledge of project languages (English/German)
• Known sensitivity to medical-grade adhesives or other skin-related complications, which might influence the outcome