Home

NoWander: Attention training for mental health

Simply put, the goal of this project is to reduce emotional suffering.

The idea came to me while on a meditation retreat: could emotional suffering (e.g. anxiety or depression) be synonymous with pathological distraction? In other words, could we plausibly frame such mental states as the inability to refocus attention away from emotionally-charged thoughts? On one hand, this seems patently obvious and little more than a restatement of the problem. However, this framing of dysphoria as repetitive self-focused attention is not new.^[1] Yet it was from this perspective that I glimpsed a potential solution: attention training.

"Attention" is, of course, an infamously vague term. This project uses it to mean "metacognition", or the awareness of the object of awareness. I believe that much emotional suffering stems from identification with thought and the inability to detect when one's mental landscape is becoming dominated by negative emotions (or possibly even the ignorance that such an ability exists). Modern times have seen the term "mindfulness" enter the zeitgeist, and our use of attention encapsulates this concept as well. However, mindfulness has become such an overloaded term that relying on it alone does little to relieve the problem. Moreover, we hope to avoid any unnecessary counter-culture pushback arising from the current "trendiness" of the mindfulness movement.

Many interventions exist for emotional disorder, and indeed mindfulness training is gaining in popularity among them. I beleve the key might lie in a closed-loop system for attention training. The team at Neuroscape recently published a paper their closed-loop medtitation app MediTrain,^[2] with promising results. The feedback which "closes the loop" in their paradigm is self-assesed user feedback. As long-term meditation practitioners can attest, however, a novice meditator is reliably misinformed about the nature of their own mind. If you ask a first-time meditator to hold attention on his breath for 10 seconds, he will try it and tell you that, yes, he can.

He is usually wrong.

The flaw of this approach is in the very definition of the problem we're trying to solve: the untrained mind is simply not aware that it is distracted. It mistakes thinking about meditation for meditation itself. It is therefore not a good source of feedback in a closed-loop system. This project aims to overcome this shortcoming with clever experimental design and machine learning, as is explained below. By learning to reliably detect distraction, we can provide a better signal to the closed-loop system and ultimately a better intervention.

This type of distraction is also known as "mind-wandering", leading to the name of this project: NoWander.

There is support for this view in the literature. In their 2006 review paper,^[3] Smallwood and Schooler frame mind-wandering as an executive control process that "can be engaged without explicit awareness" and that is "automatically initiated by a personally relevant goal." Moreover, Smallwood et al. (2007)^[4] showed a correlation between mind-wandering and dysphoria. Subjects rated as depressive displayed a higher frequency of mind-wandering, and while lost in thought they displayed both heightened physiological arousal and impaired information processing.

Roadmap

• Phase 1 - Data Collection Pipeline
  • Type: Engineering
  • Deliverable: Software to record biosignals and recoveries from a meditator
  • Status: Complete
• Phase 2a - Model Design
  • Type: Data Science
  • Deliverable: Design of mind-wandering detector model; requirements for data processing pipeline
  • Status: Complete
• Phase 2b - Single Subject
  • Type: Data Collection
  • Deliverable: 1000+ recorded recoveries from single subject
  • Status: Complete
• Phase 3 - Data Processing Pipeline
  • Type: Engineering
  • Deliverable: Software to pre-process raw recorded data for learning
  • Status: Complete
• Phase 4 - Model Training
  • Type: Data Science
  • Deliverable: Realtime mind-wandering detector
  • Status: Complete
• Phase 5 - NoWander v0.1
  • Type: Engineering
  • Deliverable: Integrate mind-wandering detector with feedback mechanism for complete closed-loop system
  • Status: Unstarted
  • ETA: TBD
• Phase 6 - Multiple Subjects
  • Type: Data Collection
  • Deliverable: 5000+ recorded recoveries from multiple subjects
  • Status: Unstarted
  • ETA: TBD
• Phase 7 - Model Validation
  • Type: Data Science
  • Deliverable: ML model tuned for generalizability and robustness
  • Status: Unstarted
  • ETA: TBD
• Phase 8 - Trials
  • Type: Clinical Trials
  • Deliverable: Studies validating intervention with outcome measures
  • Status: Unstarted
  • ETA: TBD

References

1. Pyszczynski, T., & Greenberg, J. (1987). Self-regulatory perseveration and the depressive self-focusing style: A self-awareness theory of reactive depression. Psychological Bulletin, 102(1), 122–138. https://doi.org/10.1037/0033-2909.102.1.122
2. Ziegler, D. A., Simon, A. J., Gallen, C. L., Skinner, S., Janowich, J. R., Volponi, J. J., … Gazzaley, A. (2019). Closed-loop digital meditation improves sustained attention in young adults. Nature Human Behaviour, 3(7), 746–757. doi: 10.1038/s41562-019-0611-9
3. Smallwood, J., & Schooler, J. W. (2006). The restless mind. Psychological bulletin, 132(6), 946.
4. Smallwood, J., O'Connor, R. C., Sudbery, M. V., & Obonsawin, M. (2007). Mind-wandering and dysphoria. Cognition and Emotion, 21(4), 816-842.