Abstract
Thyroid cancer is a common malignancy with broad systemic endocrine and metabolic effects (Hammond et al., 2024). Regarding diagnostics, if infra-slow body-surface potential (BSP) features reflect a systemic state, then this type of cancer is a reasonable clinical test case. Prior electrodermal work has focused on faster transients above 0.5 Hz (Boucsein, 2012; Posada-Quintero and Chon, 2020), and task-state skin potential abnormalities have been shown to distinguish mood disorders from healthy controls (Lyu et al., 2024). We concentrated on an infra-slow band from 0.025 to 0.2 Hz, called Band 1 herein. We evaluated whether infra-slow features carry reproducible group-level information when artifacts are explicitly controlled, with one pre-specified primary endpoint. "Separable" here means a group-level statistical difference, not individual-level diagnosis. In a cohort of 321 cancer patients, Band 1 energy showed a clear group-level association with thyroid cancer, which persisted after automated quality control (QC) and under stricter artifact-removal settings.