Abstract
The presence of the RUNX1::RUNX1T1 fusion gene in acute myeloid leukemia (AML) is associated with distinct morphological characteristics of myeloblasts. This study aims to assess the capability of artificial intelligence (AI) in identifying genetic abnormalities based on cell morphology. This multicenter trial included 205 patients diagnosed with AML, of which 75 were AML with RUNX1::RUNX1T1. A dataset comprising 65,039 images of myeloblasts was compiled for training, testing, and validating an AI model. The model demonstrated proficiency in adapting to varied clinical scenarios by applying two different threshold values. Under the threshold of 0.59, the testing and validation cohorts demonstrated sensitivities of 92.86% and 95.65%, with corresponding accuracies of 87.04% and 71.88%. Conversely, by setting the threshold at 0.88, specificities of 92.31% and 92.68% were achieved, along with accuracies of 88.89% and 90.63%. Regardless of the threshold, this AI model effectively distinguished RUNX1::RUNX1T1 genetic alterations based on cell morphology.