Identifying fetal status with fetal heart rate: Deep learning approach based on long convolution

CTG (Cardiotocography) is an effective tool for fetal status assessment. Clinically, doctors mainly evaluate the health of fetus by observing FHR (fetal heart rate). The rapid development of Artificial Intelligence has led realization of computer-aided CTG technology, Intelligent CTG classification based on FHR is a fundamental component of these technologies. Its implementation can provide doctors with auxiliary decisions. Most of existing FHR classification methods are based on combing different deep learning models, such as CNN (Con-volutional Neural Network), LSTM (Long short-term memory) and Transformer. However, these studies ignore the balance of positive and negative samples in dataset and the matching degree between model and FHR classification task, which reduces the classification accuracy. In this paper, we mainly discuss two major prob-lems in previous FHR classification studies: reduce class imbalance and select appropriate convolution kernel. To address above two problems, we propose a data augmentation method based on ECMN (Edge Clipping and Multiscale Noise) to resolve class imbalance. Subsequently, we introduce a one-dimensional long convolutional layer, which use trend area to calculate the appropriate convolution kernel. Based on appropriate convolution kernel, an improved residual structure with attention mechanism named TGLCN (Trend-Guided Long Convo-lution Network) is proposed to improve FHR classification accuracy. Finally, horizontal and longitudinal ex-periments show that the TGLCN obtains high classification accuracy and speed of parameter adjustment.

Automated summary

CTG (Cardiotocography) is an effective tool for assessing fetal status. Doctors assess fetal health mainly by observing FHR (fetal heart rate). This paper discusses the problems of class imbalance and appropriate convolution kernel selection in previous FHR classification studies. A data augmentation method based on ECMN is proposed to address class imbalance, and a one-dimensional long convolutional layer is introduced to calculate the appropriate convolution kernel. Furthermore, an improved residual structure with an attention mechanism called TGLCN is proposed to improve FHR classification accuracy.