The use of Artificial Intelligence and machine learning in basic research and clinical neuroscience is increasing. AI methods enable the interpretation of large multimodal datasets that can provide unbiased insights into the fundamental principles of brain function, potentially paving the way for earlier and more accurate detection of brain disorders and better informed intervention protocols. Despite AI’s ability to create accurate predictions and classifications, in most cases it lacks the ability to provide a mechanistic understanding of how inputs and outputs relate to each other. Explainable Artificial Intelligence (XAI) is a new set of techniques that attempts to provide such an understanding, here we report on some of these practical approaches. We discuss the potential value of XAI to the field of neurostimulation for both basic scientific inquiry and therapeutic purposes, as well as, outstanding questions and obstacles to the success of the XAI approach.

Electrical stimulation of the cervical vagus nerve is an emerging treatment for inflammatory bowel disease (IBD). However, side effects from cervical vagal nerve stimulation (VNS) are often reported by patients. Here we hypothesized that stimulating the vagus nerve closer to the end organ will have fewer off-target effects and will effectively reduce intestinal inflammation. Specifically, we aimed to: (i) compare off-target effects during abdominal and cervical VNS; (ii) verify that VNS levels were suprathreshold; and (iii) determine whether abdominal VNS reduces chemically-induced intestinal inflammation in rats. An electrode array was developed in-house to stimulate and record vagal neural responses. In a non-recovery experiment, stimulation-induced off-target effects were measured by implanting the cervical and abdominal vagus nerves of anaesthetized rats (n = 5) and recording changes to heart rate, respiration and blood pressure during stimulation (10 Hz; symmetric biphasic current pulse; 320 nC per phase). In a chronic experiment, the efficacy of VNS treatment was assessed by implanting an electrode array onto the abdominal vagus nerve and recording in vivo electrically-evoked neural responses during the implantation period. After 14 days, the intestine was inflamed with TNBS (2.5% 2,4,6-trinitrobenzene sulphonic acid) and rats received therapeutic VNS (n = 7; 10 Hz; 320 nC per phase; 3 h/day) or no stimulation (n = 8) for 4.5 days. Stool quality, plasma C-reactive protein and histology of the inflamed intestine were assessed. Data show that abdominal VNS had no effect (two-way RM-ANOVA: P ≥ 0.05) on cardiac, respiratory and blood pressure parameters. However, during cervical VNS heart rate decreased by 31 ± 9 beats/minute (P ≥ 0.05), respiration was inhibited and blood pressure decreased. Data addressing efficacy of VNS treatment show that electrically-evoked neural response thresholds remained stable (one-way RM ANOVA: P ≥ 0.05) and therapeutic stimulation remained above threshold. Chronically stimulated rats, compared to unstimulated rats, had improved stool quality (two-way RM ANOVA: P < 0.0001), no blood in feces (P < 0.0001), reduced plasma C-reactive protein (two-way RM ANOVA: P < 0.05) and a reduction in resident inflammatory cell populations within the intestine (Kruskal–Wallis: P < 0.05). In conclusion, abdominal VNS did not evoke off-target effects, is an effective treatment of TNBS-induced inflammation, and may be an effective treatment of IBD in humans.