Capsule Endoscopy Image Interpretation System and Method

Deep Learning-based Capsule Endoscopy Image Interpretation System

This technology describes a capsule endoscopy image interpretation system and method that preprocesses capsule endoscopy images, determines the presence of lesions using a convolutional neural network, and generates grad-CAM.

Accurately and efficiently interpreting a large volume of capsule endoscopy images has traditionally been time-consuming and prone to errors.

This technology enhances the accuracy and efficiency of interpretation by utilizing convolutional neural networks and grad-CAM to detect lesions and provide a basis for the diagnosis.

‍

‍

Key Features:

• A preprocessing unit that preprocesses capsule endoscopy images
• A convolutional neural network that determines the presence or absence of lesions
• A grad-CAM acquisition unit that provides a basis for diagnosis
• Improved interpretation accuracy and efficiency