When liver disease diagnosis encounters deep learning: Analysis, challenges, and prospects

Area of Science:

• Hepatology and Medical Imaging
• Artificial Intelligence in Medicine

Background:

• The liver, vital for metabolism and detoxification, is susceptible to damage from various factors like viruses, obesity, and alcohol, leading to liver disease.
• Traditional liver disease diagnosis relies on subjective, time-consuming clinical expertise.
• Deep learning (DL) offers potential solutions for objective and efficient liver disease diagnosis and treatment.

Purpose of the Study:

• To provide a comprehensive overview of deep learning applications in liver research.
• To analyze trends, methods, and relationships between data modalities, liver topics, and applications in 139 recent studies.
• To identify challenges and future expectations for DL in the field of hepatology.

Main Methods:

• Systematic literature review of 139 papers on deep learning in liver research published within the last five years.
• Analysis of relationships between data modalities, liver topics, and applications using Sankey diagrams.
• Summarization of deep learning methodologies applied to specific liver research topics.

Main Results:

• Identified key trends and relationships between data types, research areas, and DL applications in liver disease.
• Detailed the specific deep learning methods employed across various liver research topics.
• Highlighted the growing integration of DL in assisting with liver disease diagnosis and treatment.

Conclusions:

• Deep learning is increasingly vital for advancing liver research, offering improved diagnostic and therapeutic capabilities.
• Further research is needed to address current challenges and fully realize the potential of DL in hepatology.
• The study provides a roadmap for understanding the current landscape and future directions of AI in liver disease research.