Enhanced Blood Cancer Detection using CNN (VGG16 Model)

Abstract

This work dives into the world of deep learning software, focusing specifically on Convolutional Neural Networks (CNNs) using the VGG16 architecture. The goal? To improve the detection of blood cancer through microscopic images. Traditional diagnostic methods often struggle with accuracy and speed, making advanced techniques essential for clinical imaging. To tackle this issue, we gathered a comprehensive set of annotated blood cell images, categorizing them into routine and cancerous types. We applied data augmentation techniques to enhance the dataset and prevent overfitting during training. The VGG16 model was selected for its deep architecture, which is excellent for feature extraction, and we fine-tuned it specifically for classifying blood cell images. This involved tweaking some layers of the model to better align with the unique characteristics of our dataset. To evaluate the model's performance, we used key metrics like precision, accuracy, and F1-score. The results showed significant improvements over traditional detection methods, confirming that the model effectively identifies blood cancer cells. These findings suggest that VGG16 could serve as a reliable diagnostic tool in clinical settings, boosting the capabilities of healthcare professionals in cancer detection. This exploration highlights the groundbreaking potential of deep learning in medical diagnostics, offering a promising alternative to conventional tools. Moreover, it sets the stage for future advancements, emphasizing the need for larger and more diverse datasets, as well as collaboration with healthcare professionals for practical applications