PROJECT TITLE :

Deep Learning for Smartphone-Based Malaria Parasite Detection in Thick Blood Smears

ABSTRACT:

This study looks into the possibilities of using smartphones to detect malaria parasites in thick blood smears. We've created the first smartphone-based deep learning algorithm for detecting malaria parasites in thick blood smear photos. There are two processing steps in our method. First, we use an intensity-based Iterative Global Minimum Screening (IGMS) to locate parasite candidates by quickly screening a thick smear image. Then, using a customized Convolutional Neural Network (CNN), each candidate is classified as either a parasite or a background candidate. We offer a dataset of 1819 thick smear photographs from 150 patients openly available to the scholarly community in conjunction with this paper. As explained in this work, we used this dataset to train and test our deep learning algorithm. In terms of the following performance indicators, a patient-level five-fold cross-evaluation demonstrates the effectiveness of the customized CNN model in discriminating between positive (parasitic) and negative image patches: accuracy (93.46 percent 0.32 percent ), AUC (98.39 percent 0.18 percent ), sensitivity (92.59 percent 1.27 percent ), specificity (94.33 percent 1.25 percent ), p On both the picture and patient levels, high correlation coefficients (>0.98) between automatically discovered parasites and ground truth illustrate the applicability of our technology. Conclusion: Using deep learning algorithms, promising results were produced for parasite detection in thick blood smears for a smartphone application. Automated parasite identification on smartphones offers a promising alternative to manual parasite counting for malaria diagnosis, particularly in locations where parasitologists are scarce.