Clinical impact of the AML ML system on laboratory turnaround time

Determine the true clinical impact on laboratory turnaround time (TAT) of deploying the real-time machine-learning-based acute myeloid leukemia (AML) detection and triage system for flow cytometry within the described clinical laboratory workflow.

Background

The authors developed and clinically deployed a real-time machine-learning system that analyzes flow cytometry screening panels to detect acute myeloid leukemia (AML) and trigger additional testing steps. One intended benefit was to reduce turnaround time by accelerating triage and add-on studies.

Post-deployment analyses were confounded by concurrent process changes (e.g., digital workflow implementation, priority queue for critical cases, and autoverification), and the authors state that they could not yet determine whether the system reduces TAT and that more data over time are needed.