The application relates to the technical field of
cell image analysis and
biological cell screening, and discloses a
screening method for liver stem cells, wherein the method comprises the following steps: acquiring an original image; performing effective image screening; performing
background correction, denoising enhancement and contrast compensation; performing single-
cell instance segmentation; performing single-
cell feature extraction and trajectory association; combining a neural
network model to determine target cells; and outputting a
liver stem cell determination result. Compared with the screening mode in the prior art which relies on manual observation or single-frame
static image discrimination, especially under the condition that mature liver cells,
bile duct epithelial cells, interstitial cells and cell fragments exist simultaneously in a primary liver cell mixed suspension, the technical problem that stable and automatic identification of liver stem cells cannot be realized is solved. Since the continuous
processing flow of
image quality gating, instance segmentation and double-
branch feature fusion determination is constructed, the accuracy and
automation degree of
liver stem cell screening are improved.