A Comparative Study Using Deep Learning Model for BI-RADS Classification of Mammogram
| dc.contributor.author | Ralf Ashley anak Razali | |
| dc.date.accessioned | 2026-04-22T05:00:58Z | |
| dc.date.issued | 2025 | |
| dc.description | Breast cancer is one of the major causes of death among women all over the world, and its prognosis largely depends on early detection. Among various techniques used for screening, mammography is a popular one, but its manual interpretation generally leads to variability and diagnostic errors. This study explores the potential of deep learning models in enhancing the accuracy and efficiency of BI-RADS classification for mammograms. For that purpose, three different deep learning architectures are implemented using a publicly available dataset in Kaggle-ResNet, DenseNet, and VGG-and assessed. This preprocessing made data compatible for underrepresented BI-RADS categories to help in class balancing by cleaning, augmenting, and resizing. Accordingly, the models went through hard tests using different metrics such as accuracy, sensitivity, specificity, AUC-ROC. This opens a great opportunity to decrease diagnostic variability, support radiologists in clinical workflows, and thus contribute to the early detection of breast cancer, so important for good patient outcomes. The deployment of the best-performing model on a cloud-based demo platform underlines the practical applicability of this research, allowing stakeholders to interact with the system and assess its diagnostic capability in real-time. These findings demonstrate the transformative potential of deep learning in medical imaging and offer a stable solution for improving the diagnosis of breast cancer, thereby supporting the use of AI-powered tools in healthcare. | |
| dc.identifier.uri | https://scholarhub.unimas.my/handle/123456789/453 | |
| dc.language.iso | English | |
| dc.publisher | UNIVERSITI MALAYSIA SARAWAK | |
| dc.relation.ispartofseries | Faculty of Computer Science and Information Technology | |
| dc.subject | Breast cancer, mammography, BI-RADS classification, deep learning, DenseNet | |
| dc.title | A Comparative Study Using Deep Learning Model for BI-RADS Classification of Mammogram | |
| dc.type | Final Year Project |