Visceral pain in Irritable Bowel Syndrome (IBS) is difficult to evaluate objectively due to its complex physiological nature and lack of reliable biomarkers. Given the complexity of IBS, a multimodal physiological monitoring approach, combining electrodermal activity (EDA), electrocardiogram (ECG), and surface electromyography (sEMG), offers a promising approach to capture the autonomic and muscular responses linked to visceral pain. However, no existing wearable device supports simultaneous EDA, ECG, and sEMG acquisition from the abdomen in a format suitable for long-term, real-world use. This study presents the development and validation of a novel wearable belt for concurrent ECG, sEMG, and EDA monitoring, with EDA measured at both the torso and wrist. The system was built using modified BITalino platforms with custom-fabricated reusable electrodes and Bluetooth connectivity for real-time smartphone display. Signal quality was validated against laboratory-grade systems in 20 healthy participants during a four-stage protocol involving cognitive, orthostatic, muscular, and combined stress tasks. Time and frequency-domain analyses showed high correlations and comparable spectral features across all modalities. The belt maintained stable skin contact even during movement-intensive tasks. By enabling anatomically targeted, multimodal data acquisition, this wearable system supports real-world visceral pain assessment in IBS and is ready for deployment in ambulatory and home-based monitoring scenarios.

Alzheimer’s disease (AD), the leading cause of dementia, lacks an inexpensive and non-invasive method for early detection. Here, we investigate whether electrodermal activity (EDA), a measure of changes in electrical conductance at the skin’s surface, can offer a convenient avenue for AD screening by capturing the autonomic dysfunction associated with the disease’s pathology. As few studies have comprehensively examined EDA’s utility as a standalone biomarker for AD, we conducted a pilot study with 10 cognitively healthy controls and 20 patients with AD, evaluating EDA recordings collected during an orthostatic test and a Montreal Cognitive Assessment (MoCA). The AD group included both carriers from Colombian kindreds affected by autosomal-dominant AD and patients with sporadic late-onset AD. We compared EDA features between groups using statistical analysis and tested whether machine learning models could separate the two groups using these features. During the MoCA, phasic EDA measures, specifically the number of skin conductance responses (NSSCR) and mean time-variant sympathetic tone (TVSymp), were significantly lower in the AD group than in controls (NSSCR: $p=0.0083$, Cohen’s $d=-1.07$; mean TVSymp: $p=0.0166$, $d=-0.96$), suggesting a blunted sympathetic response to cognitive stress in AD. No significant between-group differences were observed during the orthostatic test (all $p\ge 0.60$). A random forest classifier trained on MoCA-derived EDA features achieved a balanced accuracy of 78% (95% CI 60 to 93%; 75% sensitivity, 80% specificity, AUROC 0.75), with phasic features contributing most to performance. Performance did not differ significantly from that of the support vector machine or logistic regression models. These results provide preliminary, hypothesis-generating evidence for the utility of EDA in screening for AD. The small sample size, unequal group allocation, sex imbalance, and substantial mean age disparity between groups (≈23.5 years) limit the strength of these conclusions and necessitate replication in larger, age- and sex-matched cohorts that include prodromal stages.