Identification of Sickle Cell Anemia by Employing Hybrid Optimization and Recurrent Neural Network

Red blood cell deficiencies are the root cause for anaemia. In sickle cell anaemia, the red blood cell's (RBCs) shape shifts into a crescent or sickle form. It takes a lot of time and effort to manually examine tiny images. Red blood cells have complicated and varied forms, and their overlapping cells make categorization and counting of them difficult. In this study, sickle cell identification in microscopic images is automated. The RBCs are then classified into three forms: round, elongated (sickle cell), and other shapes using image processing and deep learning approaches. Sickle cell illness fatalities can be decreased and the condition may be managed successfully with early discovery. In order to identify sickle cell disease, hybrid optimization, and recurrent neural networks have been developed. This technique first gathers images of blood cells. The image is enhanced by using the wiener filter during the initial processing step. Images of red blood cells are then segmented using Otsu thresholding. The hybrid whale and particle swarm optimization (WPSO) technique is used for extracting the characteristics from images. In order to categorize the RBC cells, a recurrent neural network (RNN) classifier is utilized. A variety of factors, including precision, accuracy, recall, and f1-score, are used to evaluate the efficiency of the suggested approach. The results of the research have been compared with other methods now in use. In comparison to current techniques, the suggested approach offered more accuracy and precision. Saving people's priceless lives would be made possible by this automated detection method.