Image understanding methods in biomedical informatics and digital imaging

Area of Science:

• Biomedical Informatics
• Cardiovascular Imaging
• Artificial Intelligence

Background:

• Coronary artery stenoses, caused by arteriosclerosis plaques, are primary contributors to ischemic cardiovascular diseases.
• Accurate analysis of coronary artery morphology is crucial for diagnosing and managing these conditions.
• Current diagnostic methods may benefit from advanced computational approaches.

Purpose of the Study:

• To explore novel applications of image recognition and AI in biomedical informatics.
• To present methods for semantically oriented analysis of 2D coronary artery images from coronography.
• To detail computer-aided recognition of coronary artery lumen stenoses using syntactic pattern recognition.

Main Methods:

• Application of image recognition and artificial intelligence (AI) in biomedical informatics.
• Semantically oriented analysis of 2D coronary artery images.
• Syntactic pattern recognition methods, specifically attribute, context-free grammar of look-ahead LR(1) type.

Main Results:

• Demonstration of new possibilities for AI and image recognition in analyzing coronary artery images.
• Successful computer analysis and recognition of local stenoses in coronary artery lumens.
• Enabling analysis of coronary artery lumen morphology through syntactic analysis.

Conclusions:

• AI and syntactic pattern recognition offer advanced capabilities for detecting coronary artery stenoses.
• These methods can improve the analysis of coronary artery morphology, aiding in the diagnosis of cardiovascular diseases.
• The presented approach holds potential for enhancing the early detection and management of arteriosclerosis-related conditions.