Fig. 7: Comparative analysis of classification performance for MCI detection models using wearable EEG.
From: Wearable EEG devices in the detection of mild cognitive impairment: a systematic review
Each chart represents one study; axes show algorithms or conditions; distance from center indicates accuracy (concentric circles: 25%, 50%, 75%, 100%). Colors distinguish experimental conditions in multi-condition studies. Highlighted labels indicate best-performing methods. Studies systematically compared 3–9 algorithms, including traditional machine learning (SVM, RF, kNN) and deep learning approaches. Results demonstrate task-dependent optimal algorithm selection, with ensemble methods and SVMs frequently achieving competitive performance. Row 1: task/algorithm comparisons; Row 2: device configurations and feature combinations; Row 3: EEG modalities and frequency bands; Row 4: encoding/decoding and emotion recognition tasks. MCI mild cognitive impairment, RF random forest, SVM support vector machine.
