Electroacupuncture modulates electroencephalographic microstate dynamics to alleviate chronic insomnia: a machine learning approach for predicting individual treatment response.

Researchers studied how electroacupuncture affects brain activity in people with chronic insomnia and whether brain measurements could predict who would respond best to treatment. The study included 41 patients with chronic insomnia and 19 healthy volunteers for comparison. Using electroencephalography (EEG) to measure brain wave patterns, scientists analyzed specific brain activity states called "microstates" before and after a 4-week course of electroacupuncture treatment.

The results showed that patients with chronic insomnia had different brain activity patterns compared to healthy sleepers, particularly in certain microstate measurements. After receiving electroacupuncture twice weekly for four weeks, patients experienced significant improvements in sleep quality and depression scores. The brain wave patterns also changed, with some measurements returning to levels similar to those of healthy sleepers. Specifically, changes in "microstate B" patterns appeared most responsive to treatment.

In an exploratory analysis using machine learning algorithms, researchers found that baseline brain measurements showed preliminary promise for predicting which patients would respond best to electroacupuncture, achieving moderate accuracy. The most helpful predictive features included specific microstate durations and transition patterns between different brain states.

For patients considering electroacupuncture for chronic insomnia, these findings suggest the treatment may help normalize brain activity patterns associated with better sleep. While the study was limited by its small sample size and needs validation in larger groups, it offers early evidence that brain measurements might someday help personalize treatment approaches. If you're considering acupuncture for insomnia, seek treatment from a licensed acupuncturist with appropriate credentials in your jurisdiction.

This study investigated EEG microstate dynamics in 41 chronic insomnia patients versus 19 healthy controls, examining changes following 4-week electroacupuncture treatment and exploring machine learning prediction of treatment response. At baseline, CI patients showed altered temporal metrics in microstates B, A, and C compared to controls. Following EA treatment, all clinical outcomes improved significantly (p<0.001), including Pittsburgh Sleep Quality Index scores. Microstate analysis revealed increased Coverage_B, Duration_B, and Occurrence_C, with multiple transition probabilities modulated post-treatment. Notably, microstate B temporal indicators normalized to healthy control levels. Treatment response (≥50% PSQI reduction) was modeled using multi-stage feature selection with eight ML algorithms; Random Forest performed optimally (AUC=0.849). Top predictive features included Duration_A, transition probabilities (D→B, C→B), Occurrence_C, and Duration_B. While preliminary and requiring validation in larger cohorts, baseline microstate parameters show exploratory potential as neurophysiological predictors for individualizing EA treatment in chronic insomnia management.