AttentionCARE: replicability of a BCI for the clinical application of augmented reality-guided EEG-based attention modification for adolescents at high risk for depression

Richard Gall, Nastasia Mcdonald, Xiaofei Huang, Anna Wears, Rebecca B. Price, Sarah Ostadabbas, Murat Akçakaya, Mary L. WoodyFrontiers in Human Neuroscience 2024 · DOI

Introduction

Depressive disorders often involve "affect-biased attention," a cognitive pattern where individuals reflexively prioritize negative emotional stimuli over goal-directed tasks. This bias is particularly significant during adolescence, a period of rapid brain development where emotional processing systems often mature faster than regulatory ones. Research suggests that daughters of depressed mothers are at a higher risk for developing depression, partly due to these attentional biases. While Attention Bias Modification Training (ABMT) has been proposed as an intervention, the field has struggled with inconsistent results and poor replicability.

The AttentionCARE Protocol

To address these challenges, researchers developed AttentionCARE, an augmented reality (AR) brain-computer interface (BCI) designed to provide real-time neurofeedback on attentional states. The system utilizes a Microsoft HoloLens AR headset integrated with a 16-channel EEG cap.

The core of the protocol involves tracking Steady-State Visual Evoked Potentials (SSVEPs). Participants are presented with two overlapping images: a task-relevant "Gabor patch" (a series of lines) and an emotional distractor (an angry or sad face). Each image flickers at a distinct frequency (8.57 Hz and 12 Hz). When a participant focuses on one of the images, their brain activity shows a corresponding spike at that specific frequency, allowing the BCI to objectively measure where their attention is directed.

The stimulus consists of an emotional face (f1) and a Gabor patch (f2) overlaid to create a composite image.

Personalized Machine Learning and Reliability

A major innovation in AttentionCARE is the use of participant-specific Support Vector Machines (SVMs). Unlike generalized algorithms, this personalized approach accounts for the high variability in EEG signals between individuals. The system "learns" a user's unique brain patterns during a brief baseline phase before providing real-time feedback.

A participant wearing the AR-EEG setup (A) and the real-time bar chart feedback (B) indicating attention levels.

The study demonstrated exceptional internal reliability, with Guttman coefficients ranging from 0.963 to 0.998. This high level of consistency addresses the "reliability crisis" often seen in previous BCI and ABMT research, suggesting that SSVEPs provide a robust "ground truth" for measuring attention in real-time.

Key Results and Clinical Potential

The study, involving 10 adolescent girls at high risk for depression, yielded promising results:

  • Classification Accuracy: The personalized SVM achieved an average accuracy of 77% in the preliminary sample and 71% in the replication sample.
  • Attentional Discrimination: Probability scores were significantly higher when participants were instructed to attend to the Gabor patch versus the emotional face, confirming the BCI's sensitivity.
  • Ecological Validity: By using AR, the training occurs within the user's actual environment, potentially making the skills more transferable to real-world situations compared to fully immersive virtual reality.

Conclusion and Future Directions

While AttentionCARE successfully measured and provided feedback on attention, the researchers noted that a single session was not enough for participants to significantly improve their attentional control. Future work will focus on multi-session "dosing" to break long-standing habits of negative bias. Additionally, moving beyond static photos to more dynamic, person-specific triggers (like social media content) could further enhance the clinical relevance of this intervention for adolescents.

Summary updated

Abstract

Affect-biased attention is the phenomenon of prioritizing attention to emotionally salient stimuli and away from goal-directed stimuli. It is thought that affect-biased attention to emotional stimuli is a driving factor in the development of depression. This effect has been well-studied in adults, but research shows that this is also true during adolescence, when the severity of depressive symptoms are correlated with the magnitude of affect-biased attention to negative emotional stimuli. Prior studies have shown that trainings to modify affect-biased attention may ameliorate depression in adults, but this research has also been stymied by concerns about reliability and replicability. This study describes a clinical application of augmented reality-guided EEG-based attention modification ("AttentionCARE") for adolescents who are at highest risk for future depressive disorders (i.e., daughters of depressed mothers). Our results (<i>n</i> = 10) indicated that the AttentionCARE protocol can reliably and accurately provide neurofeedback about adolescent attention to negative emotional distractors that detract from attention to a primary task. Through several within and cross-study replications, our work addresses concerns about the lack of reliability and reproducibility in brain-computer interface applications, offering insights for future interventions to modify affect-biased attention in high-risk adolescents.
BibTeX
@inproceedings{gall2024attentioncare,
  author = {Richard Gall and Nastasia Mcdonald and Xiaofei Huang and Anna Wears and Rebecca B. Price and Sarah Ostadabbas and Murat Akçakaya and Mary L. Woody},
  title = {{AttentionCARE: replicability of a BCI for the clinical application of augmented reality-guided EEG-based attention modification for adolescents at high risk for depression}},
  booktitle = {Frontiers in Human Neuroscience},
  year = {2024},
  doi = {10.3389/fnhum.2024.1360218},
}
Figure