Figure 2 | Scientific Reports

Figure 2

From: Comparison of different input modalities and network structures for deep learning-based seizure detection

Figure 2

The different input forms used in the present study. (a) A down-sampled raw time-series EEG with 500 data points. (b) Periodogram result with 100 data points. (c) Image of an STFT at 50 × 20 pixels. The gray STFT image was pseudo-colored for demonstration purposes. (d) Image of an EEG waveform at 40 × 250 pixels. (e) Concatenated image of three temporally separated images of EEG waveforms at 40 × 750 pixels. The images are not presented to reflect their actual sizes.

Back to article page