Bone age assessment method based on fine-grained classification

Abstract

The invention discloses a bone age evaluation method based on fine-grained image classification. The bone age evaluation method comprises the following steps: 1) acquiring an X-ray film image of a left hand wrist bone; 2) inputting a bone age evaluation network based on fine-grained classification designed in the invention to carry out bone age evaluation; and step 3) obtaining a bone age evaluation result. According to the method, the bone age of the left hand wrist bone X-ray film is evaluated by utilizing a fine-grained image classification method, a bone age evaluation network based on fine-grained image classification is designed and realized, and during bone age evaluation, the to-be-evaluated X-ray film and the gender corresponding to the X-ray film only need to be input, so that the evaluated bone age value can be obtained. For an input X-ray film, the bone age evaluation network can adaptively extract a plurality of regions of interest containing most feature information, and accurate bone age evaluation is performed by using image features of the regions of interest. The method can be used for accurately evaluating the bone age of the left hand wrist bone X-ray film of the teenagers of 0-18 years old, and has a relatively high application value.

Description

technical field [0001] The invention relates to a bone age assessment method based on fine-grained classification. Background technique [0002] Bone age assessment typically analyzes an X-ray image of the left wrist to determine an individual's bone age. The traditional bone age assessment methods are based on manual work. The reader evaluates the bone age by comparing the size and shape of the wrist bones in the X-ray to be evaluated and the standard X-ray. There are three types of traditional bone age assessment methods: counting method, atlas method and scoring method. These evaluation methods all require experts with certain professional knowledge, and the evaluation of an X-ray film often takes a long time. Thanks to the development of neural networks, more and more bone age assessment methods based on deep learning have been proposed. These methods can complete an X-ray bone age assessment within a few seconds, and the accuracy of the assessment results is comparabl...

AI Technical Summary

Problems solved by technology

Among these methods, some methods directly perform feature extraction on the entire X-ray image for bone age assessment. Although such methods use the features of the entire X-ray image, they cannot determine which parts of the X-ray image are important for bone age. Evaluation plays a key role, methods are less interpretable

Another part of the method first segments the X-ray film, extracts the regions of interest, and then uses the convolutional neural network to extract the image features of these regions of interest for bone age assessment. These methods are highly interpretable, and the selection of regions of interest is mostly based on computational The evaluation criteria of the sub-method require certain prior knowledge and additional labeling of the target region of interest

Images