Artificial intelligence-assisted multi-parameter flow cytometry gating for minimal residual disease detection in acute lymphocytic leukemia: A systematic review and meta-analysis

Abstract Background: Acute Lymphocytic Leukemia(ALL) is the most prevalent childhood cancer and a major hematologic malignancy in adults. Minimal residual disease (MRD) detection by multiparameter flow cytometry (MFC) is critical for prognosis and treatment decisions in ALL. Artificial intelligence (AI)-assisted gating has emerged as a promising approach for improved accuracy and efficiency in MRD analysis. However, the integration of AI-assisted gating into routine MRD monitoring in ALL remains limited. Objectives: The purpose of this study was to critically assess the empirical evidence of the efficacy of AI-assisted MFC gating for MRD Detection in ALL. Methods: A comprehensive and systematic electronic database search was conducted in Scopus, PubMed, Cochrane Library, and Google Scholar. Modified PICOS criteria were used to screen and select the eligible studies from the potential articles retrieved from the database search. Studies were considered if they included patients with ALL whose MRD was measured using AI-assisted approaches. The selected studies were assessed for the risk of bias using the risk of bias visualization tool (ROB 2.0), Risk of Bias in Non-randomized studies with Interventions (ROBINS-I) developed by the Cochrane Collaboration, and the National Institutes of Health (NIH) quality assessment tool for observational cohort and cross-sectional studies. Data were then procedurally extracted and analyzed. Results: The study selection process identified 11 studies, including 4373 patients. AI-assisted MFC gating demonstrated high diagnostic accuracy, with a pooled AUC of 0.954 (p < 0.001) and a high sensitivity of 0.998, effectively eliminating false-negative results that are critical for clinical decision-making. In addition, concordance analysis showed consistency with manual expert gating, with a concordance rate of 0.915. On the other hand, AI systems have significantly optimized workflows through reductions in analysis time. Moreover, AI-derived MRD negativity was significantly associated with improved overall survival and relapse-free survival. Conclusion: AI-assisted MFC gating demonstrates high diagnostic performance through high accuracy and sensitivity, while improving workflow efficiency by reducing analysis time. In addition, AI-assisted MRD detection provides standardized and reproducible analysis, essential for high-quality diagnostic capabilities.KEYWORDS: Artificial Intelligence; Multiparameter Flow Cytometry; Gating; Minimal Residual Disease; Acute Lymphocytic Leukemia.