GI Blog

AI in Capsule Endoscopy: Revolutionizing Gastrointestinal Diagnostics

07 August 2023

Capsule endoscopy has emerged as a revolutionary technique in the field of gastrointestinal diagnostics, allowing physicians to visualize the entire digestive tract with ease and precision. However, the sheer volume of images generated during a capsule endoscopy procedure can be overwhelming for healthcare professionals, leading to potential diagnostic challenges. This is where Artificial Intelligence (AI) steps in, transforming the way we analyze and interpret capsule endoscopy data.


The greatest advantage of AI in capsule endoscopy lies in its ability to analyze and interpret the vast amount of visual data generated during the procedure. AI algorithms can efficiently process thousands of images, detecting abnormalities, lesions, ulcers, and other gastrointestinal conditions that may have been overlooked by human analysis alone.


AI-powered analysis enhances the accuracy and efficiency of detecting gastrointestinal abnormalities. By comparing patterns and features in the images with vast databases of previously diagnosed cases, AI algorithms can aid in the early detection and diagnosis of various gastrointestinal disorders, including Crohn's disease, ulcerative colitis, and small intestine tumors.


Capsule endoscopy images can sometimes be challenging to interpret due to poor image quality or ambiguous findings. AI algorithms can help reduce the chances of missed diagnoses by flagging potential abnormalities, ensuring that healthcare professionals give careful attention to specific areas.


VueSmart, a cutting-edge AI software designed to revolutionize gastrointestinal diagnostics. Powered by advanced AI algorithms, VueSmart brings unprecedented efficiency and accuracy to the interpretation of capsule endoscopy data.


With its SmartScan function, capsule endoscopy is no longer a daunting task. SmartScan was trained and validated using a staggering 5,825 small bowel capsule endoscopy examinations from 51 medical centers, as published in JAMA Network Open[1]. The results were astounding, showing a 10.7% higher sensitivity and a remarkable 89.3% less reading time compared to conventional reading methods (CR).


Utilizing proprietary algorithms, SmartScan processes the raw files by deleting up to 90% redundancy, identifying and tagging as many as 16 types of abnormalities, which enables physicians to quickly and accurately identify abnormalities, lesions, and other gastrointestinal conditions.


The SmartView reading mode allows physicians to review only the images selected by SmartScan, providing a comprehensive overview of the examination. Highlighted marks above the time bar indicate specific images, which can be displayed as a short video for a clearer understanding of abnormalities.


SmartFinding takes assistance to the next level by displaying predictive results for selected images. It helps physicians quickly identify and analyze prominent characteristics, suggesting descriptive information for a more comprehensive diagnosis.


VueSmart is not just an AI software, it can be a trusted companion in delivering superior gastrointestinal care. Reading cannot be easier.


[1] Xia Xie, et al. JAMA Network Open. 2022; 5(7): e2221992. doi:10. 1001/jamanetworkopen.2022.21992 Development and Validation of an Artificial Intelligence Model for Small Bowel Capsule Endoscopy Video Review.