Lymphoma, carcinoma, and benign lesions are common diseases that have to go through several stages to be detected due to their structural similarity. A common way to distinguish these diseases is by analysing them manually based on the cell images at a certain magnification. However, this method still has many shortcomings in terms of accuracy as it is vulnerable to possible human error and requires quite a lot of time. Thus, an alternative faster and more accurate method of detection should be developed to increase patients' chances of survival. One solution to overcome this problem is by using a deep learning model which mimics the behaviour of nerve cells in the human brain and has proven to be able to classify certain diseases in many studies. As there are many existing deep learning designs, this paper aims to explore the methods of detecting these diseases and find the best performance (highest accuracy) among them. With the microscope images data (magnifications of 100 and 400 times) provided by the medical faculty of the University of Padjadjaran (UNPAD), we investigated the classification result using different kinds of deep learning designs which were our designed CNNs and transfer learning using the Inception-V3, VGG16, and MobileNet. It was found that the best model used to classify images with a magnification of 100x is MobileNet (accuracy of 53% for benign lesions and 47% for lymphoma) and designed CNN (accuracy of 59% for benign lesions and 41% for lymphoma). While for image classification with a magnification of 400x, Inception-V3 showed the best result (accuracy of 80% for carcinoma and 50% for lymphoma).
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.