Artificial intelligence-based pathology for gastrointestinal and hepatobiliary cancers

Area of Science:

• Digital pathology
• Computational oncology
• Medical imaging analysis

Background:

• Histopathology images of gastrointestinal (GI) and liver cancers are information-rich but challenging for human interpretation.
• Artificial intelligence (AI) offers advanced capabilities for analyzing complex visual data in cancer diagnostics.

Purpose of the Study:

• To explore the applications of AI in analyzing digitized histopathology slides for GI and liver cancers.
• To highlight AI's potential in assisting pathologists and improving cancer care.

Main Methods:

• Utilizing AI algorithms for the analysis of digitized histopathology slides.
• Developing AI tools for tumor detection, outcome prediction, and molecular alteration inference.

Main Results:

• AI can automatically detect tumor tissue in histopathology slides, reducing pathologist workload.
• AI demonstrates capacity to identify prognostically relevant features for outcome prediction.
• AI can infer molecular and genetic alterations directly from digital histopathology slides.

Conclusions:

• AI presents significant clinical applications in GI and liver cancer pathology.
• Pathologists and clinicians need to understand AI principles, capabilities, and limitations in pathology.
• AI-powered histopathology analysis promises to improve cancer diagnosis and patient management.