Type 2 Diabetes Mellitus Tongue Dataset
Description
This dataset comprises tongue images categorized into two groups: non-diabetic tongue images and tongue images from patients with Type 2 Diabetes Mellitus (T2DM). The dataset is designed to support the development of deep learning models for detecting T2DM based on tongue image characteristics. These images capture various features such as tongue texture, color, and coating, which are known to differ between healthy individuals and those with T2DM. The dataset can be employed for training, validating, and testing machine learning algorithms aimed at automating the early screening of Type 2 Diabetes based on tongue analysis.
Files
Steps to reproduce
Tongue images were collected using a 48 MP smartphone camera. Each participant was photographed five times to ensure dataset variability, with consistent lighting and standardized procedures to minimize external factors. For validation, medical records and blood tests were conducted to confirm the diabetes status of the diabetic group and to verify the absence of diabetes in the non-diabetic group before inclusion in the dataset.