AN IMPROVED ALGORITHM FOR PROCESSING BLOOD CELL IMAGES AND EFFICIENT DISEASE DETACTION

This paper presents an enhanced algorithm for the processing of blood cell images to improve the accuracy and efficiency of disease detection. The proposed method integrates EfficientNet as the encoder and a customized U-Net decoder for segmentation, followed by a CNNbased classification of identified cells. The algorithm addresses challenges in noisy background removal, precise segmentation, and accurate classification of different blood cell types. It was tested on publicly available datasets and compared with traditional models such as ResNet and VGG in terms of accuracy, Dice coefficient, and inference time. The results show a significant improvement in segmentation accuracy (Dice = 0.89) and disease classification performance. The proposed model provides a reliable and efficient solution for medical image analysis, offering potential applications in hematology diagnostics and automated laboratory systems.