In the inference process of existing deep learning models, it is usually necessary to process the input data level-wise, and impose a corresponding relational inductive bias on each level. This kind of relational inductive bias determines the theoretical performance upper limit of the deep learning method. In the field of sleep stage classification, only a single relational inductive bias is adopted at the same level in the mainstream methods based on deep learning. This will make the feature extraction method of deep learning incomplete and limit the performance of the method. In view of the above problems, a novel deep learning model based on hybrid relational inductive biases is proposed in this paper. It is called CCRRSleepNet. The model divides the single channel Electroencephalogram (EEG) data into three levels: frame, epoch, and sequence. It applies hybrid relational inductive biases from many aspects based on three levels. Meanwhile, multiscale atrous convolution block (MSACB) is adopted in CCRRSleepNet to learn the features of different attributes. However, in practice, the actual performance of the deep learning model depends on the nonrelational inductive biases, so a variety of matching nonrelational inductive biases are adopted in this paper to optimize CCRRSleepNet. The CCRRSleepNet is tested on the Fpz-Cz and Pz-Oz channel data of the Sleep-EDF dataset. The experimental results show that the method proposed in this paper is superior to many existing methods.

Background Infiltration is important for the surgical planning and prognosis of pituitary adenomas. Differences in preoperative diagnosis have been noted. The aim of this article is to assess the accuracy of machine learning analysis of texture-derived parameters of pituitary adenoma obtained from preoperative MRI for the prediction of high infiltration. Methods A total of 196 pituitary adenoma patients (training set: n = 176; validation set: n = 20) were enrolled in this retrospective study. In total, 4120 quantitative imaging features were extracted from CE-T1 MR images. To select the most informative features, the least absolute shrinkage and selection operator (LASSO) and variance threshold method were performed. The linear support vector machine (SVM) was used to fit the predictive model based on infiltration features. Furthermore, the receiver operating characteristic curve (ROC) was generated, and the diagnostic performance of the model was evaluated by calculating the area under the curve (AUC), accuracy, precision, recall, and F1 value. Results A variance threshold of 0.85 was used to exclude 16 features with small differences using the LASSO algorithm, and 19 optimal features were finally selected. The SVM models for predicting high infiltration yielded an AUC of 0.86 (sensitivity: 0.81, specificity 0.79) in the training set and 0.73 (sensitivity: 0.87, specificity: 0.80) in the validation set. The four evaluation indicators of the predictive model achieved good diagnostic capabilities in the training set (accuracy: 0.80, precision: 0.82, recall: 0.81, F1 score: 0.81) and independent verification set (accuracy: 0.85, precision: 0.93, recall: 0.87, F1 score: 0.90). Conclusions The radiomics model developed in this study demonstrates efficacy for the prediction of pituitary adenoma infiltration. This model could potentially aid neurosurgeons in the preoperative prediction of infiltration in PAs and contribute to the selection of ideal surgical strategies.

No abstract