<![CDATA[Procedures for diagnosis of lymphoma includes blood tests, CT scan or MRI, and histopathological examination through a biopsy. Histopathological examination is the gold standard of diagnosis. Pathology diagnosis of lymphoma is challenging and difficult in the field of diagnostic pathology. This study aims to identify lymph node metastases using the Faster R-CNN algorithm using histopathological images of lymph nodes so that the Faster RCNN system design can help the medical team to make diagnostic decisions. Identification carried out by Faster R-CNN is by classifying histopathological images into normal classes and metastatic classes. Loss values that are not indicated for underfitting and overfitting are shown from the 10th epoch to the 20th epoch. The optimizer and the number of epochs for the optimal value of 83.3% accuracy and 71.8% recall are ADAM with 20 epochs. The accuracy and recall results obtained are quite good. 1113 metastatic images and 1478 normal images were predicted correctly, while 437 metastatic images and 82 normal images were predicted incorrectly.]]>
Copyrights © 2023
