The mutation of isocitrate dehydrogenase(IDH)gene is an important biomarker for the diagnosis and prognosis of gliomas,but the current clinical detection of this biomarker relies on expensive gene sequencing technology.To address this issue,an automatic typing method for glioma IDH gene based on deep learning algorithms is proposed.Firstly,the proposed method combines convolutional neural network and Transformer to train an initial classification network using the existing glioma IDH genotyping labels and glioma magnetic resonance image data.The semi-supervised learning method is used to train the model with and without IDH gene labeled data and magnetic resonance image data.Experimental results show that the proposed method can improve the classification accuracy by 7%,surpassing other classical algorithms and the latest algorithms,providing a valuable reference for the diagnosis and treatment of glioma patients.
维普
万方数据