Abstract
Aim: In this study, we aimed to evaluate the effectiveness of artificial intelligence for the histopathological diagnosis of prostatic adenocarcinoma by analyzing the digitized pathology slides.
Materials and Methods: After the approval of the research project by the Ethics Committee of the University of Lahore - Islamabad Campus, a total of eight hundred and two (802) images were obtained from the anonymized slides stained with hematoxylin and eosin, which included 337 anonymized images of prostatic adenocarcinoma and 465 anonymized images of nodular hyperplasia of prostate. Eighty percent (80%) of the total digital images were used for training and 20% for testing. Three ResNet architectures ResNet-18, ResNet-34, and ResNet-50 were employed for the analysis of these images.
Results: in the present study, the analysis of pathology images by convolutional neural network architecture ResNet-50 has revealed the diagnostic accuracy of 993 %, while the other convolutional neural network architectures ResNet-18 and ResNet-34 showed the diagnostic accuracy of 97.1% and 98 %, respectively.
Discussion: The findings of the present study suggest that an intelligent vision system is possibly a worthwhile tool for the histopathological evaluation of prostatic tissue to differentiate between benign and malignant disorders. The application of deep learning for the histological diagnosis of malignant tumors could be quite a helpful tool for better patient care.