Skip to main content
Brain Informatics

Table 6 Improved accuracies for CS prediction and detection using 5 channel EEG trained cubes at all time segments in the entire baseline recording

From: Prediction and detection of virtual reality induced cybersickness: a spiking neural network approach using spatiotemporal EEG brain data and heart rate variability

 

EEG

EEG + ECG fusion

Prediction

85.9% (F7, 110-112s I-O connection)

74.2% (75–105 s, SNS)

Detection

76.6% (FP1, Cz I-O connection)

72.6% (2 min, SNS + SI)

  1. Top accuracies are highlighted in bold
  2. Analysis included time segment and data length optimization (prediction only), and SDSP optimization of the mod, driftup and driftdown parameter. Prediction used modified KNN algorithm. EEG trained on top 5 channels. Neuron proportions at 110-112s: O1 (9%), F8 (9%), F7 (7%), P8 (7%), T7 (6%).  Fusion accuracies increased for detection but not for prediction
Back to article page