The invention discloses a chest X-
ray image
anomaly detection method,
system and device, and relates to the field of
image processing, and the method comprises the steps: obtaining a to-be-processed chest X-
ray image; determining complete text description information according to the chest X-
ray image, a preset universal cue word and a vision-
language model, and encoding the complete text description information to obtain a first text feature;
processing the complete text description information to obtain category
label text information for describing a
disease entity, and coding the category
label text information to obtain a second text feature; extracting image features of the chest X-ray image; and according to the first text feature, the second text feature and a preset feature enhancement fusion strategy, determining a multi-
modal feature so as to determine an
anomaly detection result of the chest X-ray image according to the multi-
modal feature. According to the scheme, self-prompt text guidance can be achieved, the detection efficiency is high, the detection precision is high, and the cross-domain generalization ability is higher.