Artificial intelligence prognostication of liver disease using imaging.

Area of Science:

• Radiology
• Medical Informatics
• Hepatology

Background:

• Accurate prognostication in chronic liver disease (CLD) is crucial for patient management and cost reduction.
• Artificial intelligence (AI) offers potential for improving prognostic models through advanced image analysis.
• Current prognostic tools for CLD require enhancement for better clinical utility.

Purpose of the Study:

• To review the application of AI in prognostic models for CLD using various imaging modalities.
• To assess the strengths, weaknesses, and clinical translation challenges of AI-driven prognostic tools in CLD.
• To identify future research directions for AI in CLD prognostication.

Main Methods:

• Narrative review of AI applications in prognostic models for CLD.
• Analysis of AI integration with ultrasound (US), computed tomography (CT), and magnetic resonance imaging (MRI).
• Examination of AI-based models using a prognostic endpoint-based approach.

Main Results:

• AI algorithms show promise in segmentation, detection, and classification tasks for CLD imaging.
• AI can extract imaging features or build predictive models directly for prognostication.
• Limitations in AI application and data hinder widespread clinical translation.

Conclusions:

• AI holds significant potential to improve prognostic accuracy and clinical outcomes in CLD patients.
• Addressing current limitations is key for the successful clinical implementation of AI in CLD prognostication.
• Further research is needed to overcome challenges and optimize AI-based prognostic tools for CLD.