# MI-013: A review of classification algorithms for EEG-based brain-computer interfaces

> 本页由本地论文卡片生成。请不要直接编辑本页；修改源卡片后运行 `vp run docs:generate`。

> 中文支持说明：本页是中文站点镜像，保留论文标题、证据状态和来源字段的原始表述，避免未经核实的翻译改变证据边界。

> 内部 wiki 边界：本页提供论文卡片证据和用户提供的 PDF 下载入口，仅用于私有阅读；全文抽取文本不发布。

> 来源: `library/paper_cards/MI-013.md`

## 论文访问

* 内部 PDF: <a href={"/papers/MI-013.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)" }}>下载论文 PDF</a>
* DOI / 官方页面: [10.1088/1741-2560/4/2/R01](https://doi.org/10.1088/1741-2560/4/2/R01)
* 部署边界: 这些 PDF 链接只适合私有/受保护的内部 wiki。

## MI-013: A review of classification algorithms for EEG-based brain-computer interfaces

## Metadata

* ID: MI-013
* Title: A review of classification algorithms for EEG-based brain-computer interfaces
* Year: 2007
* DOI / URL: 10.1088/1741-2560/4/2/R01
* Local PDF: 见上方论文访问区块
* Text artifact: local-only path withheld from docs site
* Review status: `extracted`

## Study Type

* Track: MI / BCI classifiers
* Task: review EEG-BCI classification algorithms
* Participants or dataset: review paper; no single cohort
* Device/electrode setup: multiple EEG-BCI paradigms reviewed
* Protocol/task: motor imagery, movement intention, P300, cursor control, and other BCI classification contexts

## Methods

* Signal processing or analysis: taxonomy of classifiers including linear, neural-network, Bayesian, nearest-neighbor, and combinations with feature extraction
* Metrics discussed: accuracy, kappa, mutual information, sensitivity, specificity
* Online/offline: review of published BCI classifier evidence

## Key Results

* The review establishes the classical classifier landscape before deep learning.
* It emphasizes that classifier selection is only one part of a full BCI pipeline.

## Limitations

* Review stops before modern EEGNet, Riemannian transfer, and foundation/self-supervised approaches.
* Many cited performance values come from heterogeneous protocols.
* Does not itself validate SAH-BRI-Grasp.

## Relevance To Current Review

* Historical classifier baseline for MI sections.
* Helps connect CSP/FBCSP and later deep classifiers to a broader BCI classification lineage.

## Evidence Status

| Claim | Status | Evidence Note |
| --- | --- | --- |
| EEG-BCI classifier evidence must be interpreted by paradigm and protocol. | verified | The review separates classifier tables by BCI protocol/task. |
| Accuracy alone is not the only useful BCI metric. | verified | Local text lists accuracy, kappa, mutual information, sensitivity, and specificity. |
| This review proves modern SAH-BRI-Grasp decoder performance. | needs confirmation | It is historical and not system-specific. |

## Open Questions

* Which classifier families should be mentioned in the related-work survey versus omitted for focus?
* Should the paper cite both Lotte 2007 and Lotte 2018 as historical/update pair?