Medical image processing method and medical image recognition method and device

Abstract

The invention discloses a medical image processing method based on artificial intelligence. The medical image processing method comprises the steps of obtaining a to-be-identified medical image; obtaining a feature map corresponding to the to-be-identified medical image and a medical image content identification result through a medical image classification model, the feature map being obtained after convolution processing is performed on the to-be-identified medical image through a convolution layer of the medical image classification model, and the feature map having N channels; according tothe feature map and a weight parameter set, generating a thermodynamic diagram corresponding to the medical image content recognition result, wherein the weight parameter set comprises N weight parameters; and generating an image recognition result corresponding to the to-be-recognized medical image according to the thermodynamic diagram. The invention further discloses a medical image recognition method and an image recognition result display method and device. According to the invention, good interpretability is provided for the model, and a powerful basis is provided for automatic diagnosis, so that the model has more credibility, and the reliability of diagnosis based on medical images is improved.

Description

technical field

[0001] The present application relates to the field of artificial intelligence, in particular to a medical image processing method, a medical image recognition method and a device. Background technique

[0002] With the development of computer science and technology, many works have applied deep learning methods to the automatic diagnosis of medical images, which can alleviate the phenomenon of missed diagnosis and misdiagnosis caused by inexperienced doctors or excessive fatigue of doctors to a certain extent.

[0003] At present, the medical image diagnosis method based on deep learning first needs to collect a large amount of image data, which is accurately labeled by professional doctors; then use these accurately labeled images to train the deep learning model so that it can accurately fit The mapping relationship between the input image and the corresponding label. Finally, input the unlabeled original medical image into the model, and the correspondin...

Images