Intended Use Statement: Glyco is intended to be used in a non-clinical setting to analyse continuous glucose monitoring data from healthy individuals. Glyco uses statistical methods to extract useful information from glucose data. In addition, Glyco also enables visualising the glucose data to get insights from the continuous glucose monitoring time series.
Glyco is intended to be used on time series of glucose in units sepcified under units.md.
Glyco should not be used for analysing data of individuals with diabetes. Glyco is not intended to be used on individuals using insulin. Thus, Glyco should not be used to evaluate a person's risk for developing diabetes, nor to replace a patient's evaluation, nor to make, confirm or reject diagnoses.
Indications for Use: Glyco is designed for the following indications:
Analysis of continuous glucose monitoring (CGM) data in healthy individuals. Extraction of insights and patterns from CGM time series data. Visualization of glucose data for personal health monitoring and data interpretation. Device Limitations: Glyco has certain limitations:
It is not suitable for analyzing data from individuals with diabetes. It is not intended for use on individuals who are using insulin therapy. Glyco may not provide accurate results for individuals with rare or complex glucose-related conditions. Results should be interpreted in conjunction with other relevant clinical information.
Clinical Impact of Performance: The clinical impact of Glyco's performance is limited to non-clinical settings and should not be used as the sole basis for any clinical decisions. Any findings or insights generated by Glyco should be validated through clinical evaluation if intended for clinical decision-making.
Preprocessing Steps: Glyco employs the following preprocessing steps:
Data normalization to ensure consistency in glucose unit measurements. Missing data imputation using appropriate statistical techniques. Noise reduction to eliminate artifacts in the CGM data. Temporal data alignment for accurate time series analysis. Outlier detection and handling to enhance data quality
Parameters:
Final Threshold and Explanation:
Glyco utilizes a diverse set of databases containing CGM data from healthy individuals. These databases are continuously updated to improve algorithm performance and accuracy.
Glyco relies on ground truth data to validate its results. Ground truth data consists of glucose measurements from individuals with known health status. This data is essential for training and evaluating the algorithm.
Patient Population Description for FDA Validation Dataset: The FDA validation dataset includes CGM data from a representative sample of healthy individuals. The dataset covers a range of ages, genders, and demographics to ensure the algorithm's applicability to a diverse population.
Ground Truth Acquisition Methodology: The ground truth for the validation dataset is obtained through clinical assessments, medical records, and laboratory measurements. Individuals in the dataset have undergone thorough health evaluations to confirm their healthy status.
Algorithm Performance Standard: Glyco's performance will be evaluated against predefined criteria set by the FDA, including sensitivity, specificity, positive predictive value, and negative predictive value. The algorithm's performance must meet or exceed these standards to gain FDA approval for its intended use.
Please note that this document provides an overview of Glyco's healthcare considerations and algorithm description. Detailed technical documentation and validation results are available upon request.