CELLVISION: HYBRID YOLO-CNN MODEL FOR LEUKEMIA SCREENING

Abstract

Early and accurate detection of leukemia is critical for effective treatment and improved patient outcomes. Traditional diagnostic methods, while reliable, are often timeconsuming, resource-intensive, and prone to human error. To address these challenges, this paper introduces CellVision, a hybrid YOLOCNN–based model designed for automated leukocyte detection and classification in leukemia screening. By combining the realtime object detection capability of YOLO (You Only Look Once) with the deep feature extraction power of Convolutional Neural Networks (CNNs), the proposed model achieves superior accuracy in identifying and categorizing leukocytes from microscopic blood smear images. The integration of YOLO ensures fast and precise localization of leukocytes, while CNN enhances classification performance by learning discriminative morphological features. Extensive experimental evaluation on publicly available hematological image datasets demonstrates that CellVision outperforms conventional machine learning and standalone deep learning approaches in terms of precision, recall, and F1-score. The results highlight the potential of CellVision as a robust, scalable, and clinically applicable tool for assisting pathologists in leukemia diagnosis, thereby reducing diagnostic delays and supporting personalized treatment strategies.