Summary

I was contracted by the Institute for Advanced Consciousness Studies (IACS) to design a VR environment in Unity for a meditation neurofeedback system. IACS has an algorithm that predicts realtime meditative depth from the EEG of meditators. IACS wanted to turn this into a neurofeedback system. My task was to design a VR environment to visualize the meditation depth to provide neurofeedback to the meditator.

  • Researched meditation neurofeedback design

  • Distilled design principles of meditation neurofeedback systems

  • Built dynamic Unity scene with visual elements that change as a function of how deep the meditation is, and how long that depth has been achieved for

 
 

Project Description

Background

  • We have an algorithm that makes realtime predictions for how “deep” a meditation is, based on the meditator’s EEG.

  • The algorithm outputs a 33Hz distance_metric ∈ [0, 1] that measures the distance from the deepest meditative state.

Goal

  • Use visual neurofeedback in VR to help novice meditators quickly achieve deep meditative states. By seeing a visual depiction of the of their meditation, meditators can continuously monitor whether their current meditative strategy is working, and can adjust accordingly in realtime.

Spec

  • The distance_metric should be reflected visually in the VR environment in realtime.

  • After two minutes of maintaining distance_metric < 0.2, an element of the VR environment should change in a way that’s intrinsically rewarding. Four and eight min, too.

 
 

Process

  • Gather info (collect papers, collect existing VR meditation apps)

  • Read papers, try out existing VR meditation apps

  • Establish design patterns / insights

  • Brainstorm designs

  • Narrow down, get sign-off

  • Build prototype

 
 

DESIGN

The user is placed on a platform in a valley, beset by fog. They do an open awareness meditation with their eyes open. The distance_metric changes the density of the fog in the valley, updating every 30ms (subject to smoothing). After 2, 4, and 8 minutes respectively of distance_metric < 0.2, the user is rewarded by haze receding, the sun rising, and finally their perspective changing.

User is in a valley with open sky.

  • Valley creates feeling of “containment” – known to help stabilize the mind

  • Sky creates feeling of “openness”, which is an intended perceptual effect of the meditation

Distance_metric is represented as fog in valley; the fog thins as the meditation deepens.

  • Obvious visual change

  • Not distracting

  • Reveals contents within the valley – rewarding

2, 4, 8 minute rewards are given by environmental changes that induce awe and mirror perceptuals changes during meditation.

  • 2 min: haze recedes, mountains come into sharp relief

    • In open-eye meditation, meditators often encounter a perceptual increase in clarity. This is reflectec

  • 4 min: sun rises, environment gets brighter

    • perceptual reflection of brightness

  • 8 min: user rises above the valley, sees infinite ocean

    • perceptual reflection of spaciousness, awe

 
 

Design Ingredients

Principle: Design should account for the intended meditation.

Supporting Observation: A busy environment with moving animals makes it hard to sustain my concentration.

Principle: VR environment can produce emotional experiences typical of meditative experience.

Supporting Observation: I feel awe in natural VR environments. Meditation will often increase feelings of awe.

Principle: Meditators should be given clear instructions on how to relate to their environment.

Supporting Observation: In several of the VR apps I tried, it wasn’t clear how the environment related to the meditation (e.g. there were oscillating elements with no explanation).

Principle: Meditators need to acclimate to their environments.

Supporting Observation: Several studies mention including an acclimation period before the meditation starts so that users don’t feel disoriented.

Principle: Continuous motion around the user can induces quick flow states

Supporting Observation: When I tried VR meditation apps with continuous motion (e.g. an infinite runner environment, or a 360˚ particle waterfall) I quickly dropped into a flow state.

Principle: Consider tradeoffs between direct data visualization, abstract visualization, and diegetic visualization.

Supporting Observation: Some biofeedback applications visualize the biometric with a graph, whereas others visualize it with an object in the VR scene, whereas others visualize it with an abstract shape (like a circle that grows/shrinks).

Principle: There are several (not mutually exclusive) ways the VR environment can relate to the meditation.

  1. VR environment is the meditation object (e.g., “notice the impermanence of the ocean’s waves”)

  2. VR environment modulates mental state of meditator (e.g. closed space calms down the meditator)

  3. VR environment mimics perceptual features of meditating (e.g. widening FoV) 

  4. VR environment contains “sensory extension” or “sensory substitute” (e.g. breath is visualized in the environment)

  5. VR environment produces flow state (e.g. a runner game)

Supporting Observation: Different VR apps relate to the meditation itself in different ways; sometimes the user is instructed to pay attention to a breath pacer, sometimes the environment is just a background without any explicit relationship to the meditation, sometimes the environment itself induces the meditative state (e.g. flow state from continuous motion), and so on.

 
 

Next Steps

  • Open Questions

    • What impact do the changing visual elements have on people’s meditation?

      • Experiment to run: Run sessions in VR headset equipped with eye tracking; analyze eye tracking data, EEG data (meditative depth), and post-session interviews with users.

    • How should users be acclimated to the VR environment? The literature indicates this is important in order for users to feel safe, which supports their meditation.

      • Experiment to run: Compare no acclimation vs 3-minute acclimation period, interview users after.

    • Should users be told they’re being rewarded (2, 4, and 8 minute environmental changes)?

      • Experiment to run: A-B test.

    • How accurate do people perceive the metric to be, and how does that perceived accuracy impact their experience?

      • Experiment to run: Ask users about their perception of how accurate the metric is, and how that impacted their experience.

    • Should the fog density change continuously, or should it have several discrete levels?

      • Experiment to run: Compare continuous vs. discrete, looking at EEG data (meditative depth) and qualitative subject reports.