Leveraging Artificial Intelligence for Enhanced Platelet Management in Dengue Fever

Ajit Pal Singh *

Department of Medical Lab Technology, SSAHS, Sharda University, Gr. Noida, U.P, India.

Riya Pandey

Institute of Paramedical Sciences, GIMS, Greater Noida, U.P. India.

Rahul Saxena

Department of Biochemistry, SSAHS, Sharda University, Gr. Noida, U.P, India.

Suyash Saxena

Department of Biochemistry, SSAHS, Sharda University, Gr. Noida, U.P, India.

*Author to whom correspondence should be addressed.


Abstract

Dengue fever poses a considerable therapeutic challenge owing to its erratic course and the potential for severe thrombocytopenia necessitating prompt care. Utilizing artificial intelligence (AI) provides a revolutionary method for platelet management by facilitating early forecasting of the platelet's lowest point, recovery trajectories, and transfusion requirements. This study delineates a comprehensive AI-driven architecture that incorporates clinical characteristics such as age, fever duration, haematocrit levels, and white blood cell trends into supervised learning models, including Random Forest, XGBoost, and LSTM. These models, incorporated into real-time clinical decision support tools, are intended for seamless integration with hospital information systems and deliver actionable alarms for significant platelet reductions. The methodology prioritizes transparency, data confidentiality, and fair model efficacy across diverse demographics. This AI-enabled solution, through prospective validation in dengue-endemic locations and synchronization with national health systems, aims to enhance transfusion procedures, alleviate hospital strain, and markedly improve patient outcomes during dengue outbreaks.

Keywords: Dengue, platelet, fever, transfusion, strategy, artificial intelligence


How to Cite

Singh, Ajit Pal, Riya Pandey, Rahul Saxena, and Suyash Saxena. 2025. “Leveraging Artificial Intelligence for Enhanced Platelet Management in Dengue Fever”. Asian Journal of Current Research 10 (3):93-107. https://doi.org/10.56557/ajocr/2025/v10i39506.

Downloads

Download data is not yet available.