Klasifikasi Citra Sel Acute Lymphoblastic Leukemia menggunakan Jaringan Konvolusi dengan Transfer Learning Arsitektur ResNet 18

Acute Lymphoblastic Leukemia (ALL) is a type of blood cancer that commonly affects children. Early diagnosis of ALL is crucial for effective treatment. In this study, we propose a deep learning approach using convolutional neural network (CNN) with ResNet18 architecture for classifying ALL cell images. The objective of this research is to evaluate the performance of the proposed model and compare it with the results obtained from the main reference paper. ResNet18 model was chosen due to its excellent precision and utilization of fewer parameters. The dataset used in this study is "Acute lymphoblastic leukemia-Image DataBase (IDB) 2" obtained from the ALL - IDB website. The model was trained using transfer learning technique, where the last layer of the ResNet18 model was replaced with a new fully connected layer, followed by ReLU activation function, and adjusted the final output layer. The experimental results show that the proposed ResNet18 model performs exceptionally well for binary classification, achieving 100% accuracy, precision, sensitivity, and specificity. For multiclass classification, the ResNet18 model achieved an accuracy score of 88%, precision of 81.25%, sensitivity of 81.75%, and specificity of 93.5%.