Aligning computational pathology with clinical practice for colorectal cancer

Area of Science:

• Computational pathology
• Digital diagnostics
• Colorectal cancer (CRC) research

Background:

• Colorectal cancer (CRC) pathology reports contain over 20 essential elements per International Collaboration on Cancer Reporting (ICCR) guidelines.
• Current computational tools for CRC diagnostics are limited, often lacking clinical validation and focusing on few report elements.

Purpose of the Study:

• To review the current state of computational pathology in CRC diagnostics.
• To identify gaps in existing computational tools and their clinical validation.
• To guide the alignment of computational pathology with clinical practice.

Main Methods:

• Literature review of computational tools for colorectal cancer pathology reports.
• Analysis of adherence to International Collaboration on Cancer Reporting (ICCR) guidelines.
• Identification of current limitations and clinical validation challenges.

Main Results:

• Existing computational tools for CRC diagnostics are not widely clinically validated.
• Most tools focus on a limited subset of the more than 20 required pathology report elements.
• Significant gaps exist in aligning computational pathology tools with clinical needs.

Conclusions:

• There is a need for more clinically validated computational tools in colorectal cancer pathology.
• Future tools should address a broader range of pathology report elements.
• Bridging the gap between computational pathology and clinical practice is crucial for advancing CRC diagnostics.