Artificial intelligence in echocardiography: trends, hotspots and future directions

Area of Science:

• Cardiology
• Medical Imaging
• Artificial Intelligence

Background:

• Artificial intelligence (AI) is transforming echocardiography, enhancing diagnostic accuracy and automating workflows.
• Operator-dependent variability in echocardiography is a key challenge addressed by AI integration.
• Systematic analyses of AI's impact on echocardiography are currently limited.

Purpose of the Study:

• To systematically analyze the bibliometric trends and research landscape of AI in echocardiography.
• To identify key contributors, institutions, and countries in AI-echocardiography research.
• To map the evolution of research themes and keywords in this field.

Main Methods:

• Bibliometric analysis of 1,296 publications from the Web of Science database (2019-2024).
• Utilized bibliometrix R package for data analysis, including annual trends, author/institutional/country contributions, and keyword evolution.
• Examined publication output, citation trends, and thematic shifts over time.

Main Results:

• Rapid growth in publications from 45 (2019) to 302 (2024), with peak citations in 2019.
• United States and China led in publication output; Mayo Clinic and Harvard University were leading institutions.
• Key research areas included AI-driven cardiac parameter quantification, pathology detection, diagnosis, classification, and heart failure.
• Thematic evolution showed a shift from structural abnormality studies to deep learning applications.

Conclusions:

• AI integration in echocardiography is rapidly advancing, evidenced by accelerated publication rates.
• Research hotspots include automated image analysis, diagnostic classification, prognostic prediction, and workflow optimization.
• Deep learning models show promise in reducing operator dependency and enhancing diagnostic precision.
• Future research priorities are data standardization, fully automated diagnostic pipelines, and improved clinical reliability of AI systems.