In this paper, we propose a convolutional neural network (CNN) model that can effectively detect the cycle of breath sounds in COVID-19 patients. Data used in this work are collected from hospital patients, the Mel-spectrogram features were extracted from the data, the convolutional neural network is then used for training. The result shows that the sensitivity of this method is 90.03%, and the average accuracy is 91.32%.By our knowledge this work is the first one that a neural network method has been used to detect the breath cycle of COVID-19 patients.