Simulated Type 1 Diabetes Patient Data for 10 Days

Muñoz-Organero's dataset contains simulated physiological data for Type 1 Diabetes Mellitus (T1DM) patients. The data was generated using the AIDA diabetes simulator to produce 10 days of records for validating a deep physiological model for blood glucose prediction. The dataset was harvested by e-cienciaDatos and last updated on May 5, 2024.

Use Cases

• Validate blood glucose prediction models based on simulated physiological data.
• Benchmark deep learning algorithms for time-series forecasting in diabetes management.
• Train models for personalized insulin dosing recommendations based on simulated patient trajectories.
• Analyze the performance of physiological models under controlled simulation conditions.

Strengths

• Data is generated by the established AIDA diabetes simulator, providing a controlled validation environment.
• Covers a 10-day period, which may be sufficient for short-term prediction model validation.
• Dataset is associated with a peer-reviewed publication in Sensors (Switzerland), 2020.

Limitations

• Row count, column definitions, and file formats are unknown, hindering immediate usability assessment.
• Data is simulated, not real-world clinical data, which may limit generalizability to actual patient scenarios.
• Description metadata is limited; actual data quality requires manual inspection after download.

Provenance

Source

Muñoz Organero, Mario via e-cienciaDatos Harvested Dataverse

Collection Method

Generated using the AIDA diabetes simulator.

Time Range

Simulated data for a 10-day period.

Freshness

Last updated 2024-05-05 06:13:42; freshness should be verified.

Geography

Simulated data; no specific geographic coverage.