Pathology Made Simple: ChatGPT's Summarization of Pathology Reports

Area of Science:

• Medical Informatics
• Artificial Intelligence in Medicine
• Pathology Reporting

Background:

• Pathology reports are crucial for clinical decision-making but are often lengthy and complex.
• Effective communication of pathology findings is vital for patient care.
• Artificial intelligence (AI) presents potential solutions for distilling complex medical information.

Purpose of the Study:

• To evaluate the performance of ChatGPT-4o in generating single-sentence summaries of pathology reports.
• To assess if these summaries retain critical clinical information and are accessible to medical professionals.
• To determine the accuracy and readability of AI-generated pathology report summaries.

Main Methods:

• Retrospective analysis of 120 anonymized pathology reports (2022-2023).
• Focus on 40 complex cases across breast pathology, melanocytic lesions, and lymphomas.
• ChatGPT-4o generated 1-sentence summaries using a standardized prompt; two pathologists reviewed summaries for accuracy and readability.

Main Results:

• High scores for inclusion of essential information across subspecialties (8.09-9.55).
• Critical error-free rates varied by subspecialty (62.5%-95%).
• Readability was rated excellent (10/10) in 119 of 120 cases.

Conclusions:

• ChatGPT-4o demonstrates potential for generating concise and readable pathology report summaries.
• AI-assisted summarization shows promise for enhancing clinical communication and efficiency.
• Further refinement and domain-specific training are needed to improve AI reliability for complex cases.