Classifying Microscopic Images of Reactive Lymphocytosis Using Two-Step Tandem AI Models

The practical applications of automatic recognition and categorization technology for next-generation systems are desired in the clinical laboratory. We approached identification reactive lymphocytosis using artificial intelligence (AI) studied its usefulness blood smear screening. This study created one- two-step AI models lymphocytosis. ResNet-101 model was applied deep learning. original image set supervised training consisted 5765 typical nucleated cell images. subjects assessment were 25 healthy cases, erythroblast cases. total accuracy (mean ± standard deviation) 0.971 0.047 0.977 0.024 healthy, 0.938 0.040 0.978 0.018 erythroblast, 0.856 0.056 0.863 0.069 respectively. showed a sensitivity 0.960 specificity 1.000 between As our tandem high performance identifying screening, we plan to apply this method development differentiate neoplastic