Abstract
Introduction: Electroconvulsive therapy (ECT) remains one of the most effective approaches for treatment-resistant depressive episodes, despite the potential cognitive impairment associated with this treatment. As a potent stimulator of neuroplasticity, ECT might normalize aberrant depression-related brain function via the brain's reconstruction by forming new neural connections. Multiple lines of evidence have demonstrated that functional connectivity (FC) changes are reliable indicators of antidepressant efficacy and cognitive changes from static and dynamic perspectives. However, no previous studies have directly ascertained whether and how different aspects of FC provide complementary information in terms of neuroimaging-based prediction of clinical outcomes.
Methods: In this study, we implemented a fully automated independent component analysis framework to an ECT dataset with subjects (n = 50, age = 65.54 ± 8.92) randomized to three treatment amplitudes (600, 700, or 800 milliamperes [mA]). We extracted the static functional network connectivity (sFNC) and dynamic FNC (dFNC) features and employed a partial least square regression to build predictive models for antidepressant outcomes and cognitive changes.
Results: We found that both antidepressant outcomes and memory changes can be robustly predicted by the changes in sFNC (permutation test p < 5.0 × 10<sup>-3</sup>). More interestingly, by adding dFNC information, the model achieved higher accuracy for predicting changes in the Hamilton Depression Rating Scale 24-item (HDRS<sub>24</sub>, t = 9.6434, p = 1.5 × 10<sup>-21</sup>). The predictive maps of clinical outcomes show a weakly negative correlation, indicating that the ECT-induced antidepressant outcomes and cognitive changes might be associated with different functional brain neuroplasticity.
Discussion: The overall results reveal that dynamic FC is not redundant but reflects mechanisms of ECT that cannot be captured by its static counterpart, especially for the prediction of antidepressant efficacy. Tracking the predictive signatures of static and dynamic FC will help maximize antidepressant outcomes and cognitive safety with individualized ECT dosing.
Introduction
Electroconvulsive Therapy (ECT) is considered a "gold standard" for treatment-resistant depression. Despite its efficacy, it remains controversial due to potential cognitive side effects, particularly memory loss. While it is known that ECT stimulates neuroplasticity—the brain's ability to rewire itself—clinicians lack precise biomarkers to predict which patients will benefit most and who is at the highest risk for cognitive decline.
Previous neuroimaging research has focused on Functional Connectivity (FC), which measures how different brain regions communicate. Most studies have utilized static FC (sFNC), which averages brain activity over several minutes. This study addresses whether dynamic FC (dFNC)—the second-to-second fluctuations in these connections—offers complementary, non-redundant information that can improve the prediction of clinical outcomes.
Methods
The researchers analyzed a dataset of 50 patients (average age ~65) undergoing ECT at the University of New Mexico. The analysis utilized the Neuromark framework, an automated Independent Component Analysis (ICA) pipeline that identifies 53 consistent Intrinsic Connectivity Networks (ICNs) across subjects. These templates were derived from large-scale reference datasets, including the Brain Genomics Superstruct Project (GSP) and the Human Connectome Project (HCP).
Two types of features were extracted:
- Static FNC (sFNC): Calculated using Pearson correlations between the time-courses of the 53 networks, representing overall connection strength.
- Dynamic FNC (dFNC): Employed a "sliding window" approach to quantify how these correlations changed over time. This was measured through four metrics: the number of distinct "meta-states" visited, switching frequency, the range of states, and the total functional "distance" traveled.
Predictive modeling was performed using Partial Least Square Regression (PLSR) within a 10-fold cross-validation framework to predict changes in the Hamilton Depression Rating Scale (HDRS) and the Hopkins Verbal Learning Test (HVLT).
Key Results
The study demonstrated that brain dynamics provide unique information for predicting antidepressant response. While static features alone predicted HDRS changes with a correlation of $r \approx 0.52$, the inclusion of dynamic features significantly boosted this accuracy to $r \approx 0.54$ ($p = 1.5 \times 10^{-21}$).
In contrast, memory changes (HVLT) were robustly predicted by static features ($r \approx 0.50$), but adding dynamic information did not yield significant improvements. This suggests that while antidepressant efficacy is linked to the fluidity of brain states, cognitive side effects may be more related to stable, structural-like changes in connectivity. These results remained robust even when controlling for age, sex, and ECT dosage.
Discussion and Insights
A critical finding of this research is the divergent neuroplasticity underlying different outcomes. The predictive maps for depression relief and cognitive changes showed a weakly negative correlation, suggesting that the "beneficial" rewiring for mood and the "adverse" rewiring affecting memory are distinct biological processes occurring in different circuits.
Key anatomical insights included:
- The Cerebellum: Emerged as a vital player in predicting both emotional and cognitive outcomes, supporting theories that the cerebro-cerebellar pathway is essential for mood regulation.
- Default Mode Network (DMN): Decreased connectivity within the DMN—the brain's "internal monologue" circuit—was highly predictive of better antidepressant outcomes.
While the sample size of 50 is substantial for this type of study, the authors acknowledge that larger, multi-site studies are needed to move toward "individualized ECT dosing." Future work will explore non-linear deep learning models and broader cognitive domains beyond memory to further refine these predictive signatures.
Summary updated