ABO-BTI: An Open-Source ABO Blood Typing Image Dataset for Medical AI Applications

Abstract

Accurate blood type classification is cru- cial for safe transfusions and clinical decision-making, yet existing research is limited by the lack of stan- dardized, publicly available datasets for training and evaluating machine learning models. To address this gap, we introduce ABO-BTI (ABO Blood Typing Im- age), the first open-source dataset dedicated to blood type classification using high-resolution agglutination images. The dataset comprises 144 cases, with 432 images standardized to a resolution of 1280×590 pix- els after processing. This study evaluates the effective- ness of deep learning for blood type identification us- ing the ABO-BTI database. Three models, ResNet50, MobileNetV2, and a proposed deep learning architec- ture, were trained and tested on the dataset to as- sess its suitability for machine learning applications. The proposed model achieved an accuracy of 96.51%, significantly outperforming MobileNetV2 (12.64%) and ResNet50 (72.41%). Comparative analysis with tradi- tional machine learning methods further demonstrated that deep learning provides competitive performance while reducing reliance on handcrafted feature extrac- tion. These results highlight ABO-BTI as a valuable benchmark for advancing AI-driven blood type classifi- cation. The findings also suggest the potential integra- tion of deep learning-based classification into embed- ded systems for real-time blood typing in point of care and emergency settings. By providing a standardized dataset and demonstrating the viability of deep learn- ing models, this study lays the foundation for future re- search in automated blood classification, with implica- tions for both clinical applications and AI-driven med- ical diagnostics.