Single-lead Thigh ECG Dataset (tOLIet) with Analysis of BMI Effects on Cardiac Signal Quality

In previous work, we introduced an 'invisible' ECG system with electrodes integrated into a toilet seat, capturing signals from the thighs. Here, we present the tOLIet dataset with single-lead thigh ECGs to advance cardiovascular assessment using this novel approach. The dataset includes 149 records from 86 individuals (50 females, 36 males; mean age 31.73 ± 13.11 years; weight 66.89 ± 10.70 kg; height 166.82 ± 6.07 cm). Participants were recruited via the Centro Hospitalar Universitário de Lisboa Central (CHULC). Each recording features four differential signals from toilet-seat electrodes alongside reference data from a hospital-grade 12-lead ECG. Beyond signal collection and quality evaluation, we conducted a gender-specific analysis comparing valid signal percentages relative to Body Mass Index (BMI). This analysis explores anatomical or physiological factors affecting thigh-based ECG acquisition, guiding system design and customization to enhance signal reliability across populations.