International testing and refinement of AI algorithms predicting acute leukemia subtypes from routine laboratory data.
Turki AT., Fan Y., Hernández-Sánchez A., Silva W., Fleming S., Yalcin K., Van Elssen CHMJ., Madanat Y., Karasek M., Aljurf M., Della Porta MG., Martinez-Roca A., Guarnera L., Steffen K., Antoniou E., Rivas MM., Mishra DK., Blum AT., Manantsoa SN., Adiat A., Enshaei A., Thol F., Voso MT., Chen J., Elhassan TA., Moorman AV., Vidriales MB., Neuendorff NR., Koc A., Mishra P., Strumberg D., Fourmanov RS., Heine L., Kleesiek J., Munárriz D., Asti G., Mokoonlall M., Kometas M., Rego E., Mecklenbrauck R., Sobas M., Wu D., Nensa F., Engelke M.
Despite advances for patients with acute leukemia health disparities limit access to diagnosis and treatment. Artificial Intelligence (AI) approaches may address some disparities. We retrospectively assemble a diverse, international cohort of 6206 leukemia patients from 20 centers to test an AI tool designed to support leukemia diagnosis using standard laboratory results. Executing the pretrained algorithm results in varying accuracy metrics. With confidence cutoff predictions, 2000-fold bootstrapped area under the curve (AUROC) metrics are 0.94 for acute myeloid leukemia (AML), 0.98 for the promyelocytic subtype and 0.84 for acute lymphoblastic leukemia. However, this cutoff excludes 70.8-92.5% of patients from predictions. We improve accuracy and robustness, while maintaining generalizability via an ensemble of Isolation Forest and Local Outlier Factor increasing AUROC for AML from 0.72 to 0.84 (hold-out test set, patients below confidence threshold), while excluding only 12.1% of patients. Furthermore, we retrain the algorithm for pediatric patients.