Early Detection of Blood Cancer Using Deep Learning and Real-Time Web Deployment

– Early detection of blood cancer, particularly leukemia, is critical for improving patient survival rates. Traditional diagnostic techniques such as manual microscopic examination are time-consuming and highly dependent on expert interpretation. This paper presents a deep learning-based approach for automated detection of blood cancer using microscopic blood smear images. A Convolutional Neural Network (CNN) model based on transfer learning with MobileNetV2 was developed and trained on preprocessed datasets. Preprocessing techniques included resizing, normalization, stain normalization, and data augmentation. The model achieved an accuracy of 89.25% with strong precision and recall performance. Unlike conventional studies, this work extends beyond model development by integrating the trained model into a real-time web-based application. The deployed system allows users to upload images and receive instant predictions, enhancing accessibility and practical usability. This study demonstrates the potential of combining deep learning with deployable systems for early medical diagnosis.