Identification of eupneic breathing using machine learning.

The diaphragm muscle (DIAm) is the primary inspiratory muscle in mammals. In awake animals, considerable heterogeneity in the electromyographic (EMG) activity of the DIAm reflects varied ventilatory and nonventilatory behaviors. Experiments in awake animals are an essential component to understanding the neuromotor control of breathing, which has especially begun to be appreciated within the last decade. However, insofar as the intent is to study the control of breathing, it is paramount to identify DIAm EMG activity that in fact reflects breathing. Current strategies for doing so in a reproducible, reliable, and efficient fashion are lacking. In the present article, we evaluated DIAm EMG from awake animals using hierarchical clustering across four-dimensional feature space to classify eupneic breathing. Our model, which can be implemented with automated threshold of the clustering dendrogram, successfully identified eupneic breathing with high F1 score (0.92), specificity (0.70), and accuracy (0.88), suggesting that it is a robust and reliable tool for investigating the neural control of breathing. The heterogeneity of diaphragm muscle (DIAm) activity in awake animals reflects real motor behavior diversity but makes assessments of eupneic breathing challenging. The present article uses an unsupervised machine learning model to identify eupneic breathing amidst a deluge of different DIAm electromyography (EMG) burst patterns in awake rats. This technique offers a scalable and reliable tool that improves efficiency of DIAm EMG analysis and minimizes potential sources of bias.