ASC committee review maps LLM applications, validation requirements and regulatory pathways for cytopathology
The American Society of Cytopathology Clinical Practice Committee published a practical review examining the current evidence and future applications of large language models and vision-language models in cytopathology. Investigators outline use cases ranging from structured reporting and diagnostic assistance to quality control and education, while clearly distinguishing between preliminary evidence and hypothetical applications. The review highlights critical deployment barriers including hallucination risks, limited explainability, algorithmic bias, data privacy concerns, and infrastructure gaps, alongside an overview of current US and EU regulatory frameworks for AI software as a medical device. The authors recommend establishing cytopathology-specific benchmarks, pursuing multi-institutional validation, implementing transparent governance, and adopting an incremental deployment strategy starting with low-risk tasks. This guidance provides a structured framework for laboratory directors and pathologists navigating the responsible integration of generative AI into diagnostic workflows.
The original study
Leveraging large language models to enhance cytopathology: Opportunities, challenges, and future directions; a practical review from the ASC Clinical Practice Committee.
- Authors
- Bilal KH, Gibson J, Kim D, Levi AW, Nassar A, Oen H, et al.
- Journal
- Cancer cytopathology
- Type
- Journal Article, Review
- PMID
- 42585107
Original abstract
Large language models (LLMs) and vision-language models represent a fundamentally different category of artificial intelligence (AI) compared to prior image analysis approaches in digital pathology, which have largely been based on convolutional neural network architectures. This review from the American Society of Cytopathology Clinical Practice Committee examines the current evidence for LLM and vision-language model applications in cytopathology, including structured reporting, diagnostic assistance, quality control, education, and workflow integration. The distinction between applications with preliminary evidence and those that remain hypothetical is described. A detailed assessment of the challenges that must be addressed before clinical deployment, including hallucination risk, limited explainability, bias, data privacy, validation gaps, and infrastructure barriers is discussed. A review of the regulatory landscape in the United States and European Union as it applies to AI-enabled software as a medical device is provided. Recommendations addressing cytopathology-specific benchmarks, multi-institutional validation, transparent governance, and incremental deployment beginning with low-risk applications are suggested. In the current environment, LLMs have the potential to augment cytopathology practice, but responsible adoption requires rigorous validation and sustained collaboration among cytopathologists, AI researchers, and regulatory bodies.