Advancing Malaria Diagnosis Through Deep Learning: A Comparative Evaluation of Convolutional Neural Networks

Mr.B.Rajesh1 , Gaddam Hemanth2 , Gandham Mounika Sri3 , Jonnalagadda Litheesh4 , Guvvala Lokesh5
1Assistant Professor, Dept of Information Technology, SV College of Engineering, Tirupati, India.
2B. Tech, Dept of Information Technology, SV college of Engineering, Tirupati, India.
3B. Tech, Dept of Information Technology, SV college of Engineering, Tirupati, India.
4B. Tech, Dept of Information Technology, SV college of Engineering, Tirupati, India.
5B. Tech, Dept of Information Technology, SV college of Engineering, Tirupati, India.
Email: 1rajesh.b1@svce.edu.in, 2gaddamhemanthyadav@gmail.com, 3mounikasree2004@gmail.com,
4jonnalagaddalithish@gmail.com, 5dangerloki2325@gmail.com
Corresponding Author *: Mr.B.Rajesh

Abstract-Malaria is a serious infectious disease caused by Plasmodium parasites, transmitted to humans through the bite of infected female Anopheles mosquitoes. After entering the bloodstream, the parasites travel to the liver to mature and multiply, causing symptoms such as fever, chills, and fatigue within 10–15 days. If untreated, it can lead to severe complications like brain swelling, organ failure, and death. The existing detection system uses CNN architectures, including a basic CNN and VGG16, with preprocessing, regularization,and dataaugmentation applied to over 27,000 blood smear images, achieving strong performance but limited robustness under varied clinical conditions. The proposed system aims to enhance malaria detection accuracy by exploring advanced neural network architectures beyond traditional CNNs. This includes investigating Capsule Networks, ResNet, EfficientNet, and Vision Transformers to improve computational efficiency and feature extraction. Additionally, the inclusion of multimodal data is designed to advance the generalizability and interpretability of image-based diagnostics. The balance between model complexity and resource availability will be maintained through optimization of hyperparameters and learning techniques for scalable deployment in healthcare settings. This novel system presents a comprehensive AI diagnostic framework that enhances technical efficacy and practicality for effective malaria prevention