Unmarked white blood cell three-classification method based on bright field microscopic imaging

• Main Authors: WEI MAOYU, HUANG XIWEI
• Format: Patent
• Language: Chinese; English
• Published: 12.04.2022
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; HANDLING RECORD CARRIERS; PHYSICS; PRESENTATION OF DATA; RECOGNITION OF DATA; RECORD CARRIERS

The invention discloses an unmarked white blood cell three-classification method based on bright field microscopic imaging. The classification method comprises the following specific steps: 1, constructing a leukocyte data set with cell category labels; and 2, constructing a batch-free normalized label-free leukocyte classification network. And 3, training the classification network constructed in the step 2 by using the leukocyte data set constructed in the step 1. And 4, identifying the types of the leukocytes by using the trained classification network. Compared with an existing classification method based on staining, the method for classifying the unmarked white blood cells has the advantages that the pretreatment steps of the sample are simpler and more convenient, and meanwhile, the influence of a staining agent on cell viability and physiological status is overcome. In addition, an autonomously designed network structure is used, and three classifications of unmarked white blood cells are achieved. Co