Skip to main content
Fig. 4 | Journal of NeuroEngineering and Rehabilitation

Fig. 4

From: Automated calibration of somatosensory stimulation using reinforcement learning

Fig. 4

RL-based platform adapts to impaired nerves of peripheral neuropathic subjects. The results obtained by performing the mapping of the three nerves (i.e., peroneal, tibial and sural) on two neuropathic subjects (S1 and S2) using the AI mapping platform are reported. A Results of the characterization in terms of injected charge for the low and high level, location and type of evoked sensations. B RL performance in terms of average and standard deviation over the three nerves reporting time, number of stimulations delivered, injected charge (low- and high-level calibration) and quality of the mapping for each subject. Bar plots in the shaded area represents the average values of each metric for people affected by polyneuropathy (\(\overline{P }\)) and healthy participants (\(\overline{H }\)) for easier comparison (p < 0.05 (*), p < 0.01(**), p < 0.001 (***))

Back to article page