Normal bone marrow cell classification method based on medical hyperspectral microscopic image and deep learning system

The invention relates to the technical field of bone marrow cell classification, and discloses a normal bone marrow cell classification method based on a medical hyperspectral microscopic image and a deep learning system, and the method comprises the following steps: preparing a bone marrow smear; dyeing the bone marrow smear; collecting hyperspectral data of the cells by adopting a hyperspectral camera; performing training learning on the hyperspectral data, and establishing a cell classification model; and classifying the bone marrow cells through the cell classification model obtained through training. According to the method, a hyperspectral technology is combined with a deep analysis model to perform morphological analysis on various cells of the bone marrow smear, so that the accuracy of qualitative and quantitative analysis on normal and abnormal bone marrow cells can be effectively improved, and a basis is provided for applying the combination of the hyperspectral technology and an AI analysis system