Classification of Microscopic Malaria Parasitized Images Using Deep Learning Feature Fusion
Muhammad Asim · Lahore Garrison University Research Journal of Computer Science and Information Technology · 2023
An infectious disease that causes a chronic and potentially life-threatening infection causedby microorganisms of the Plasmodium class is malaria or malarial disease. It is critical to detect thepresence of Malaria parasites as early as possible to ensure that antimalarial treatment is adequate tocure the particular type of Plasmodium. This is to reduce death rates and to focus on various infectionsin the event of an adverse outcome. This study aimed to develop an artificial intelligence approachcapable of separating parasitized erythrocytes from normal basophilic erythrocytes as well as plateletsoverlying the red blood cells to overcome the high cost of Malaria diagnostic equipment. The tone andtexture characteristics of erythrocyte images were extracted using histogram thresholds and watershedmethods and then fused with Squeeze Net and ShuffleNet algorithms. The measures includedplanning, preparing, approving, and testing Deep Convolution Neural Network Segmentation withoutpreparation using a graphic processor unit. A total of 96% accuracy and specificity was obtained forthe position of malaria in red blood cells based on the overall results. It has been demonstrated thatdeep learning techniques can be effective in clinical pathology. This provides new directions fordevelopment as well as increasing awareness of researchers in the medical field. The proposed methodologyand application of fusion techniques on the images produce good results for the classificationand digital identification of the disease.