# MI-029: Continuous Three-Dimensional Control of a Virtual Helicopter Using a Motor Imagery Based Brain-Computer Interface

> Generated from a local paper card. Do not edit this page directly; edit the source card and rerun `vp run docs:generate`.

> Internal wiki boundary: paper-card evidence and user-provided PDF downloads are available for private reading. Extracted full-paper text is not published.

> Source: `library/paper_cards/MI-029.md`

## Paper Access

* Internal PDF: <a href={"/papers/MI-029.pdf"} download style={{ display: "inline-flex", alignItems: "center", justifyContent: "center", minHeight: "2.25rem", padding: "0.45rem 0.8rem", borderRadius: "6px", backgroundColor: "#047857", color: "#ffffff", fontWeight: 700, lineHeight: 1, textDecoration: "none", boxShadow: "0 1px 2px rgba(15, 23, 42, 0.22)" }}>Download Paper</a>
* DOI / official page: [10.1371/journal.pone.0026322](https://doi.org/10.1371/journal.pone.0026322)
* Open-access page: [Open access page](https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0026322)
* Deployment boundary: these PDF links are intended only for a private/protected internal wiki.

## MI-029: Continuous Three-Dimensional Control of a Virtual Helicopter Using a Motor Imagery Based Brain-Computer Interface

## Metadata

* ID: MI-029
* Title: Continuous Three-Dimensional Control of a Virtual Helicopter Using a Motor Imagery Based Brain-Computer Interface
* Year: 2011
* DOI / URL: 10.1371/journal.pone.0026322
* Local PDF: see Paper Access section above
* Text artifact: local-only path withheld from docs site
* Review status: `extracted`

## Study Type

* Track: MI
* Task: continuous 3D virtual helicopter control using sensorimotor rhythms
* Participants or dataset: three healthy subjects
* Device/electrode setup: Neuroscan SynAmps2 with BCI2000; detailed montage needs extraction
* Protocol/task: sequential training ending in 3D virtual helicopter ring-navigation task

## Methods

* Signal processing or analysis: temporal and spatial filtering of raw EEG into control signals
* Training/calibration: sequential training from simpler tasks to pseudo-3D and 3D helicopter control
* Online/offline: online virtual-control task

## Key Results

* Three subjects learned to modulate SMRs for multidimensional virtual control.
* The paper is an important comparator for strong claims about noninvasive MI continuous control.

## Limitations

* Virtual helicopter task, not physical robotic arm grasping.
* Only three trained healthy subjects.
* Training burden and task-specific adaptation are substantial.

## Relevance To Current Review

* Useful boundary paper: noninvasive MI can support impressive continuous virtual control in selected/trained settings, but this should not be generalized to robust 6-DoF arm control.
* Supports the project choice to scope MI mainly to intervention/mode control.

## Evidence Status

| Claim | Status | Evidence Note |
| --- | --- | --- |
| Noninvasive MI has been used for continuous 3D virtual control in a small trained-subject study. | verified | Abstract and methods describe three subjects controlling a virtual helicopter. |
| This is not direct evidence for physical robotic grasping. | inferred | The task is virtual and uses selected training conditions. |
| MI should replace shared autonomy for robotic-arm low-level control. | needs confirmation | The paper does not evaluate physical manipulation or grasp safety. |

## Open Questions

* Should this be cited as an upper-bound comparator or as a cautionary boundary?
* How much training did each participant need before 3D control?