Exploring the influencing factors of initial seizure time during modified electroconvulsive therapy in patients under 30 years old with mental disorders.

Objective: To investigate the factors influencing the initial seizure duration during modified electroconvulsive therapy (MECT) in patients under 30 years old with mental disorders and propose optimized initial treatment parameters for this population. Methods: A total of 287 patients were included. Six features-preset energy percentage, etomidate dose, gender, age, antiepileptic drug usage, and disease type (depression, bipolar disorder, schizophrenia)-were used to build a classification model predicting whether seizure duration was ideal (15-75 seconds) or non-ideal. Data preprocessing included one-hot encoding for categorical variables and Z-score normalization for continuous variables. Synthetic Minority Over-sampling Technique (SMOTE) was applied to balance the dataset. A random forest classification model with five-fold cross-validation was used to analyze feature importance. Results: The model achieved average precision of 76.9%, recall of 75.1%, and F1-score of 74.4%. Age (importance coefficient: 0.338), energy percentage (0.252), and etomidate dose (0.208) were the core influencing factors, collectively contributing >79% of explanatory power. Disease type, antiepileptic drug usage, and gender had minimal impacts. Conclusion: For patients under 30 years old, age, energy percentage, and etomidate dose were the most critical predictors of ideal seizure time during initial MECT, while disease type, antiepileptic drug use, and gender had minimal impact. We recommend setting the initial energy percentage to 1/3-2/5 of the patient's age, combined with etomidate dose adjustment, to effectively optimize seizure duration.