Use of deep learning techniques to classify acoustic emission data from knee joints

We report here on the application of deep learning techniques to distinguish acoustic emission signals from human knees, recorded on 30 volunteers divided into three age groups 18-34, 35-49 and 50+. The deep learning model developed and applied to this data was able to correctly identify samples from each category with success rates of up to 89.5% in some cases, an outcome that is much higher than would be expected from random selection.