High Synchrony Is Not Whole-Brain Coordination: A Model-Free Metric to Distinguish Functional from Pathological Brain States

Brain-wide synchrony is one of the most widely used descriptors of brain state. High synchrony is often interpreted as whole-brain integration. However, pathological synchrony — paroxysmal discharges in disorders of consciousness (DOC), seizure onset, burst-suppression — can produce equal or higher synchrony values without functional coordination. Existing metrics cannot numerically separate these two types of high synchrony. We introduce coverage, a model-free metric defined as the fraction of EEG channels whose power spectrum contains a bimodal structure (two frequency peaks separated by a valley) within a specified band. Coverage is orthogonal to synchrony: it measures spatial extent of bimodal structure, not temporal correlation. Across four datasets (24 meditation subjects, 3 DOC patients, 20 anesthesia patients, 20 subjects in a replication dataset), we show: High synchrony occurs in two qualitatively distinct types. A DOC patient (0201) has sync = 0.947 — higher than the meditation subject sub-017 (0.885) — but coverage = 0.000. Only coverage distinguishes functional coordination from pathological hypersynchrony. Coverage is not a proxy for signal quality. In the same subject, same session, coverage differs 3-fold between two tasks (sub-014: 11/64 vs 33/64). Coverage is not a proxy for consciousness level. Under anesthesia, theta, sync, and r_late decrease while coverage remains unchanged. These results establish coverage as the first numerically separable dimension for distinguishing functional whole-brain coordination from pathological synchrony.