Conclusions

The FETAL spectral-divergence score - the full-spectrum Jensen-Shannon divergence between a tracing’s normalized power spectrum and a synthetic reference spectrum - is an interpretable, single-valued feature whose computational behavior is now characterized. In a prespecified simulation, it remained near chance under the null condition, detected progressively larger imposed within-band spectral shifts, and degraded under artifact stress, while the scalar comparators remained near chance. These findings establish computational identifiability under the stated synthetic assumptions and support locked patient-level validation against paired umbilical-artery blood gases. They do not establish diagnostic performance, incremental predictive value, clinical utility, or improved outcomes in humans.