Detection of malaria-infected cells using deep learning approach
Archana Nandibewoor, Bhagyashri G. Bhat, Misba Sawar, Prarthana Nayak, Raksha Udupi, Abhilash Hegde · BLDE University Journal of Health Sciences · 2025
Abstract: BACKGROUND: Recent technological advancements have ushered in a new era of robust research capabilities and intricate problem-solving potential, enabling precise disease diagnoses. Among the myriad health challenges faced, malaria stands as a formidable adversary. The conventional approach to malaria diagnosis involves human technicians meticulously examining blood smears for infected red blood cells under a microscope. This process is not only time-intensive but also reliant on the expertise of the examiner, leaving room for potential human errors. AIM AND OBJECTIVE: This paper endeavors to harness the power of deep learning algorithms to automate the detection of malaria-infected cells in blood smears. MATERIALS AND METHODS: The proposed model leverages a deep learning approach, autonomously classifying infected cells and making predictions based on standard microscope slides featuring thin blood smears. RESULTS: Extensive testing and validation of the model encompass a wide array of pretrained attributes, yielding an impressive accuracy rate exceeding 50%. CONCLUSION: In an age characterized by data-driven solutions, this research strives to revolutionize malaria diagnosis, offering a sophisticated, efficient alternative to the traditional, labor-intensive methods. Through this, the healthcare sector can potentially benefit from a more streamlined and reliable approach to combating this life-threatening disease.