Enhancing Precision in Blood Cell Cancer Classification: Cutting-Edge CNN Techniques for Improved Accuracy
Eshika Jain, Pratham Kaushik · 2024
The early detection of blood cell cancer is crucial for its effective treatment and the patient's outcome. The conventional techniques employed for diagnosis are based largely on microscopic analyses of blood smears by pathologists, which are found to be time-consuming and not insignificant in terms of errors. Considering the need for an effective diagnostic tool, the current study is aimed at the development of convolutional neural networks in blood cell cancer detection. A CNN was trained on an image dataset of blood cells to identify cells containing cancer and those that do not, and achieved 98.00% accuracy, wherein precision and recall also had an identical 98.00% score. These figures indicate the identifying ability of the cells while outdated positives and negatives are reduced. In light of all this, the presented research on CNNs can dramatically aid in enhancing the state of the art of imaging in terms of precision and efficiency. The model presented can help doctors diagnose accurately by automating this process; this should improve patient care and outcomes. This supports other studies that have opined for the adoption of advanced machine learning concepts in diagnostics that provide a means to allow for early identification and treatment of blood cell cancer.