Regulatory categorization and implementation pathway for ML-based flow cytometry tools

Ascertain how machine-learning-based tools for flow cytometry, such as the described AML detection and triage system, will be categorized by the U.S. Food and Drug Administration and delineate the regulatory pathway required for clinical implementation under the current Laboratory Developed Tests (LDT) framework and related guidance.

Background

The paper situates the deployed AML detection system within a shifting regulatory environment following the FDA’s final rule on laboratory developed tests and recent guidance suggesting AI tools may require FDA clearance.

The authors explicitly note uncertainty about how such AI tools will be categorized and the consequent pathway to clinical implementation, indicating a regulatory question that remains unresolved.

References

While the FDA has issued guidance documents (Bogdanoski et al., 2024) which suggest that all AI tools must seek FDA clearance, the categorization of these tools remains unclear and their pathway to clinical implementation will require reassessment as regulations become more defined.

— Clinical Validation of a Real-Time Machine Learning-based System for the Detection of Acute Myeloid Leukemia by Flow Cytometry  (2409.11350 - Zuromski et al., 2024) in Discussion, paragraph beginning with the regulatory landscape and FDA LDT rule