Skip to main content
. 2022 May 10;17(5):e0267452. doi: 10.1371/journal.pone.0267452

Fig 8. General usability of DeepGANnel for 2-dimensional physiological time series data.

Fig 8

In this example we fed rodent ECG data (dimension 1) along with a beat annotation (dimension 2, peak of r-wave) into DeepGANnel and trained with 10,000 epochs until realistic rodent ECG data was generated. Data collected via electrocardiogram transmitters from male Wistar rats (ETA-F20; Data Sciences International, St Paul, MN, USA) as previously described [43], briefly ECG signal was digitized to a PC with a CED Micro1401 using Spike2 at 5 kHz. In principle this could also be encoded with further annotations such as t-wave (note rats do not show a significant p-wave).