Predicting Online Behavioural Responses to Transcranial Direct Current Stimulation in Stroke Patients with Anomia

Area of Science:

• Neuroscience
• Neurolinguistics
• Rehabilitation Science

Background:

• Anomia, a common language impairment after brain injury, typically requires intensive therapy.
• Neurostimulation is explored to enhance anomia treatment, but evidence is inconsistent.
• Individual responses to neurostimulation may follow predictable patterns.

Purpose of the Study:

• To predict individual behavioral responses to anodal neurostimulation in chronic anomia.
• To determine if pre-stimulation clinical, behavioral, and MRI data can predict neurostimulation effects.
• To investigate if neurostimulation effects are systematic and predictable at the individual level.

Main Methods:

• Partial Least Squares regression was used to predict neurostimulation effects.
• 36 chronic anomia patients performed naming and size judgment tasks.
• Models were trained on clinical, behavioral, and MRI data, and validated using cross-validation.

Main Results:

• Predictive models based on pre-stimulation data significantly outperformed a null model.
• Neurostimulation effects on accuracy and reaction time were predictable for both naming and size judgment tasks.
• Both performance improvements and declines were equally predictable, suggesting non-placebo effects.

Conclusions:

• Individual responses to neurostimulation in anomia are systematic and predictable.
• The predictable nature of both positive and negative effects suggests neurostimulation's impact is not solely due to placebo.
• Neurostimulation may improve or impair task performance in patients with anomia.