A Health-Based Deep Learning System for Rapid and Precise Detection of Acute Lymphoblastic Leukemia

Acute Lymphoblastic Leukemia (ALL) is a prevalent and life-threatening form of cancer that requires accurate and timely diagnosis for effective treatment. Traditional diagnostic methods for ALL often involve time-consuming and subjective manual examination of blood smears, leading to potential errors and delays in diagnosis. To address these challenges, this project proposes a diagnostic system based on deep learning Convolutional Neural Networks (CNNs) and Streamlit, aimed at achieving fast and accurate classification of Acute Lymphoblastic Leukemia (ALL). The project leverages the power of deep learning CNNs to automatically learn and extract relevant features from microscopic images of blood smears. A Leukemia dataset of annotated blood smear images, consisting of Benign, Early, Precancerous and Prognosis samples, is collected and pre-processed. The images are resized,normalized, and augmented to enhance the robustness and diversity of the training data. The proposed system utilizes the popular CNN architecture VGG16 as the backbone for feature extraction. The pre-trained weights of the CNN model, learned from largescale image datasets, are utilized to initialize the model. The final layers of the CNN are modified to suit the multi-class classification task of distinguishing between Benign, Early, Precancerous and Prognosis samples. To provide an intuitive and user-friendly interface, the Streamlit framework is employed to develop the diagnostic system. The system allows users, including medical professionals, to upload blood smear images and obtain immediate predictions on the presence of Acute Lymphoblastic Leukemia. Overall this paper presents a novel and efficient diagnostic system for the classification of Acute Lymphoblastic Leukemia using deep learning CNNs and Streamlit. The integration of advanced deep learning techniques with a user-friendly interface has the potential to revolutionize the diagnostic process, enabling timely and accurate identification of ALL.