THE JOURNAL OF JAPAN SOCIETY FOR CLINICAL ANESTHESIA
Online ISSN : 1349-9149
Print ISSN : 0285-4945
ISSN-L : 0285-4945
Symposium (1)
Application of AI Image Analysis Technology to Anesthesiology:Developing a Model to Predict Difficult Intubation and Problems
Tatsuya HAYASAKA, Yu ONODERA, Hiroaki TOYAMA
Author information
JOURNAL FREE ACCESS

2025 Volume 45 Issue 6 Pages 586-590

Details
Abstract

This article discusses the application of AI image analysis techniques in anesthesiology. We provide an overview of deep learning technologies, particularly convolutional neural networks(CNN), and explore the practical potential of image recognition AI in anesthesia practice. Drawing from our experience in developing an AI model to predict difficult intubation, we demonstrate the utility of advanced techniques such as transfer learning and Grad-CAM analysis while introducing clinical applications such as AI-assisted ultrasound-guided nerve blocks. AI serves as an “assistant” that complements the visual decision-making of anesthesiologists. By addressing these challenges, including multicenter validation and clarification of the scope of application, AI technology is expected to improve anesthesia care quality.

Content from these authors
© by The Japan Society for Clinical Anesthesia
Previous article Next article
feedback
Top