Artificial Intelligence-Assisted Whole Slide Image Analysis for Lymph Node Status Prediction in Early Colorectal and...

Area of Science:

• Gastroenterology
• Pathology
• Artificial Intelligence

Background:

• Endoscopic resection is a primary treatment for early colorectal cancer (T1 CRC) and early gastric cancer (EGC).
• A significant risk of lymph node metastasis (LNM) necessitates further surgical intervention in high-risk cases.
• Current pathological evaluation for LNM risk suffers from accuracy and reproducibility issues due to interobserver variability.

Purpose of the Study:

• To review the challenges in predicting LNM for T1 CRC and EGC.
• To explore the potential of AI-assisted whole slide image (WSI) analysis as a solution for improving LNM risk stratification.
• To discuss the current limitations and future directions for AI in this field.

Main Methods:

• Review of current literature on LNM prediction in T1 CRC and EGC.
• Analysis of studies investigating AI-assisted WSI analysis for LNM detection.
• Discussion of interobserver variability in pathological assessment.

Main Results:

• AI-assisted WSI analysis demonstrates encouraging results for LNM prediction in T1 CRC and EGC.
• AI offers a potential pathologist-independent method to enhance diagnostic consistency.
• Current AI models face limitations such as small sample sizes and insufficient external validation.

Conclusions:

• AI-assisted WSI analysis holds promise for improving LNM risk stratification in early GI cancers.
• Further high-quality evidence and validation are required for clinical implementation.
• Addressing technical challenges like stain standardization is crucial for AI adoption.