Accurate gastrointestinal (GI) lesions classification from endoscopic images is crucial in patient diagnosis. The scarcity of annotated training data limits the full clinical impact of supervised deep learning (DL) in clinical settings, which is usually limited due to the scarcity of annotated training data. To alleviate this effect of small data size, we propose a GI lesions classification method based on supervised contrastive representative learning, which creates representations of the lesion from many unannotated endoscopic images. We used 12,147 endoscopic images drawn from private and public sources. Data augmentation techniques are implemented with the encoder network and projector. Supervised contrastive loss is utilized as a loss hypermeter, and the final classification task is performed in the last phase. Our method achieves better lesion classification accuracy (96.4%) than another related state of the methods (self-supervised=94.6%, and cross entropy=93.8%). Future work will improve the robustness of the proposed method's automatic classification accuracy to detect lesion severity levels and implement the proposed approach in multi-modal medical imaging.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.