The invention provides a digital slice
abnormal cell detection method and device,
electronic equipment and a storage medium, and the method comprises the steps: carrying out the pixel-level instance segmentation of a to-be-detected
image based on a trained instance segmentation model, and obtaining a
cell mask and a
cell nucleus mask in the to-be-detected image; based on the spatial characteristics and / or attribute characteristics of the
cell mask and the
cell nucleus mask, carrying out cell-
cell nucleus association matching to obtain a cell-
cell nucleus pair; and calculating the characteristic value of the
cell morphology quantitative characteristic of the cell-cell
nucleus pair, and determining the cell with the characteristic value exceeding a characteristic threshold as an
abnormal cell. According to the digital section
abnormal cell detection method and device, the
electronic equipment and the storage medium provided by the invention, the accuracy, the
automation level and the clinical applicability of abnormal cell detection are remarkably improved through a progressive process of instance segmentation-association matching-quantitative judgment, and a more reliable and efficient intelligent tool is provided for early screening of
cervical cancer.