We present a Convolutional Neural Network (CNN) suitable to assess unconstrained surgical cuttings, trained on a dataset created with a data augmentation technique.