Expert-Labeled Clinical Dataset for Anemia Detection

Published: 30 January 2026| Version 2 | DOI: 10.17632/8sycpbhzn7.2
Contributors:
MD Hasan Ahmad, Sajib Bormon,

Description

This dataset contains 364 clinically collected samples comprising anemia and non-anemia cases, obtained from patients at Atapara General Hospital. The data were collected under real clinical conditions and reflect routine diagnostic practices in a hospital setting. All samples were carefully observed, verified, and labeled by an expert physician, Dr. Mohammad Mohiuddin Alam, ensuring the reliability and medical correctness of the annotations. The labeling process followed standard clinical assessment criteria, minimizing noise and misclassification in the dataset. The dataset includes two primary classes: • Anemia • Non-anemia Each sample represents an individual patient record and was anonymized prior to inclusion to protect patient privacy. No personally identifiable information is included in the dataset, and ethical considerations were strictly maintained during data collection and preparation.

Files

Institutions

Daffodil International University

Categories

Machine Learning, Anemia, Binary Classification

Licence