Figure 3
From: DELMEP: a deep learning algorithm for automated annotation of motor evoked potential latencies
(a) Pre-processing of the two MEPs shown in Fig. 1, divided into: trimming (1), smoothing (2), centering (3) and normalizing (4). Panels (b) and (c) show the MEPs before and after the pre-processing, respectively. When comparing (b) and (c), note that the different MEPs look similar after the pre-processing, thus facilitating the training and later use of the neural network.
