Research On Ensemble Of Multi-scale Fine-Tuning Convolutional Neural Network For Recognition Of Benign And Malignant Thyroid Nodules

In the related medical diagnosis of thyroid nodular diseases,the benign and malignant analysis of thyroid nodules is very important.Aiming at the problem of poor classification performance due to the small number of training image samples and ignoring multi-scale structure and texture information,in order to improve the accuracy of diagnosis of benign and malignant thyroid nodule,this paper proposes a method for thyroid nodule ultrasound image recognition based on ensemble of multi-scale finetuning convolutional neural networks.The main research contents and work of this paper are as follows:(1)This paper describes the related content of thyroid ultrasonic image diagnosis technology,analyzes the status quo of thyroid nodule ultrasonic image recognition technology,and expounds the significance of thyroid nodule ultrasonic image recognition research.The classification methods of thyroid ultrasonic images based on traditional machine learning,deep learning and transfer learning are compared,and the technical characteristics and shortcomings of each method are analyzed,which provides a theoretical basis for the proposed algorithm of thyroid nodule ultrasound image recognition based on ensemble of multi-scale fine-tuning convolutional neural networks.(2)The steps of constructing thyroid nodule ultrasound image recognition based on ensemble of multi-scale fine-tuning convolutional neural networks are described in detail.First,the image is converted into three different scales of information as input data,so that the model can learn the feature information of different scales of the image,and improve the feature extraction ability of the model.Secondly,nine fine-tuning models of three different scales were constructed by optimizing the full-connection layer structure of three kinds of pre-training models(Alex Net,VGG16 and Res Net50)and the transfer learning and fine-tuning strategy,so that the model could better learn the characteristic differences of target domain(thyroid ultrasound image)and source domain(Image Net).Finally,the optimal fine-tuning model combination is selected and the final integration model is obtained by the weighted fusion method of model output category probability,and the classification performance is further improved by utilizing the diversity of models.(3)Experiments are designed on the real collected data set to test and analyze the effectiveness of the algorithm in this paper,and to conduct comparative experiments with other algorithms.The accuracy,sensitivity,specificity and area under curve(AUC)of benign and malignant thyroid nodules were 96.0%,94.1%,97.7% and 0.98.The experimental results show that the algorithm is superior to the traditional machine learning algorithm and other algorithms in the field of benign and malignant thyroid nodule identification,and can effectively extract complementary visual feature information with satisfactory classification performance.