Real-Time Blood Donor Prediction and Campaign Optimization using DNN-Autoencoder and MERN Architecture

The current Blood Donation Management System (BDMS) faces several challenges, including data redundancy, inefficient donor-donation matching, delayed responses in emergency situations, and a lack of real-time updates. These issues are often exacerbated by manual management, leading to increased mismanagement of blood bank data, slow decision-making processes, and poor coordination between donors, recipients, and organizers. Additionally, the absence of a reliable notification system and a lack of security in handling sensitive data further exacerbate the difficulties in managing blood donations efficiently. This paper proposes a solution to these challenges by implementing a hybrid Deep Neural Network (DNN) and Autoencoder model integrated into a MERN stack-based system. The proposed methodology aims to automate donor registration, blood group matching, and real-time scheduling of blood donation camps, significantly reducing the time and effort needed to manage the donation process. By leveraging MongoDB’s NoSQL database, the system ensures scalability and efficient data handling, while the use of Express.js and Node.js guarantees rapid server-side processing. React.js provides a user-friendly interface that facilitates seamless interaction between donors and organizers. Key features of this work include a comprehensive dashboard for organizers to manage blood donation camps, real-time notifications for donors about upcoming camps and urgent needs, and a robust security mechanism with multi-layer authentication. Additionally, the hybrid DNN-Autoencoder model enhances the prediction accuracy for donor availability and optimizes campaign participation, improving the overall efficiency of the BDMS.