Artificial Intelligence in Chronic Lymphocytic Leukemia: From Prognostic Biomarkers to Individualized Treatment Selection

Available artificial intelligence (AI) applications in chronic lymphocytic leukemia (CLL) mainly address diagnosis, prognosis, treatment need, or measurable residual disease rather than comparative treatment selection. This state-of-the-art review examines the gap and proposes a CLL-specific clinical decision-support architecture. A focused evidence map was assembled by title-and-abstract screening of 245 unique PubMed records retrieved using a documented 2014–September 2026 search strategy. Fifty-five original CLL-relevant AI and machine learning (ML) reports were retained for descriptive mapping. Existing tools are heterogeneous and rarely establish comparative causal effects under competing regimens. We define individualized selection using clinically eligible treatment sets, explicit treatment contrasts, regimen-specific outcomes, calibration and uncertainty checks, and post hoc longitudinal patient-preference filtering. The architecture is conceptual, not a trained or externally validated treatment-selection algorithm; prospective studies will be necessary to demonstrate clinical utility.