Have a personal or library account? Click to login
Deep learning in public health: evaluating anemia detection methods Cover

Abstract

Anemia is considered as prevalent public health concern that mostly impacts children, expectant or recently gave birth mothers, adolescent females, and women going through menstruation. This work focuses on detecting anemia and classifying anemia and its type by blood cell image using machine learning models—the VGG16, InceptionV3, and DenseNet121 for early identification of anemia. The F1-score, recall, accuracy, and precision of the models were assessed after they were trained on a dataset of closer to 1000 images of blood cells. The accuracy obtained was 93.43% for the VGG16 model, 90.48% for DenseNet121, and 78.80% for the InceptionV3 model. The different evaluation metrics reveal the extent of success of each model at blood cell classification for early identification and detection of anemia and its type. The present work aims to experimentally establish machine-learning models as valuable instruments for the early identification and detection of anemia.

Language: English
Submitted on: Aug 23, 2025
|
Published on: Feb 17, 2026
In partnership with: Paradigm Publishing Services
Publication frequency: 1 issue per year

© 2026 Neelu Jyothi Ahuja, Jyoti Upadhyay, Tanupriya Choudhury, Ashish Jain, Bhupesh Kumar Dewangan, Ketan Kotecha, published by Professor Subhas Chandra Mukhopadhyay
This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License.