Chinese diabetes datasets for data-driven machine learning

Area of Science:

• Endocrinology and Metabolism
• Medical Informatics
• Data Science in Healthcare

Background:

• Effective diabetes management relies on comprehensive patient data, particularly for data-driven machine learning approaches.
• Publicly accessible datasets are crucial for advancing research in diabetes mellitus (DM).

Purpose of the Study:

• To introduce and describe the Shanghai Type 1 Diabetes Mellitus (T1DM) and Shanghai Type 2 Diabetes Mellitus (T2DM) datasets.
• To make these datasets publicly available to facilitate research in diabetes management.

Main Methods:

• Data collected under real-life conditions from Type 1 (n=12) and Type 2 (n=100) diabetic patients in Shanghai, China.
• Datasets include clinical characteristics, laboratory measurements, and medication information.
• Continuous glucose monitoring (CGM) data (3-14 days) and daily dietary information are provided.

Main Results:

• The ShanghaiT1DM and ShanghaiT2DM datasets are now publicly available for research.
• The datasets offer rich, real-world data for diabetes management studies.
• Comprehensive data includes clinical, biochemical, medication, glucose monitoring, and dietary information.

Conclusions:

• The ShanghaiT1DM and ShanghaiT2DM datasets represent a significant contribution to diabetes research resources.
• These datasets will support the development of novel data-driven algorithms and technologies for diabetes monitoring and management.
• Facilitating research through public data release is key to improving diabetes care outcomes.