Treatment-resistant depression (TRD) affects approximately 30% of major depressive disorder(MDD) cases and represents a major unmet clinical need. Ketamine produces rapid antidepressanteffects in TRD, but response is variable and no validated biomarker predicts who will respond.Multiple independent studies have now shown that baseline EEG features -- particularly vigilancestage distribution and spectral dynamics -- predict ketamine response, but no unifying mechanisticaccount of why baseline brain state should predict response to an NMDA antagonist has beenproposed. We present the Thalamic Filter Model (TFM) as a candidate mechanistic account. TheTFM proposes that depression may represent a state of thalamic over-filtering: chronicallyelevated thalamic reticular nucleus (TRN) inhibitory tone raises the thalamic impedance gate(Phi_th), narrowing conscious bandwidth and producing the cognitive rigidity, rumination, andaffective narrowing characteristic of depression. In this framework, ketamine's rapidantidepressant effect may reflect indirect TRN disinhibition via glutamatergic synapticpotentiation, transiently lowering Phi_th and expanding conscious bandwidth. Baseline EEGtemporal dynamics -- specifically lag-1 autocorrelation (AR1) and vigilance stage distribution --index individual thalamic filter state: patients with higher baseline filter impedance (lowervigilance, higher AR1) may have more room for ketamine-induced filter opening and thus greaterantidepressant response. We review published evidence from six independent ketamine EEGbiomarker studies (total n > 200) showing that lower baseline vigilance, lower baseline gammapower, and higher alpha power all predict better ketamine response -- all consistent with the TFMprediction that higher baseline filter impedance predicts greater response to filter-openingintervention. We derive three falsifiable predictions distinguishing TFM from alternative accountsand propose AR1 as a practical, low-cost baseline biomarker for ketamine response prediction.Keywords: ketamine; treatment-resistant depression; EEG biomarker; thalamic filter; thalamicreticular nucleus; AR1; autocorrelation; vigilance; antidepressant response prediction; thalamicimpedance