Right now it's a simple user-defined fixed threshold, which could obviously be improved. Thoughts right now:
1). Data-definted threshold, e.g. SD threshold
2). Right now the spike time is defined as the time of the peak (max for rising, min for falling, or max(abs) for bidirectional) after the threshold crossing, but what if the user wants the time of the threshold crossing instead?
3). Similarly, the amplitude is just the raw data value at the peak, but what about e.g. peak-to-peak amplitude (you can kind of hack this right now by using the baseline-subtracted traces but even that's not quite right).
4). Currently spikes are detected over the whole trace - should be able to define a detection window.
johnmbarrett
commented
6 years ago
Right now it's a simple user-defined fixed threshold, which could obviously be improved. Thoughts right now:
1). Data-definted threshold, e.g. SD threshold 2). Right now the spike time is defined as the time of the peak (max for rising, min for falling, or max(abs) for bidirectional) after the threshold crossing, but what if the user wants the time of the threshold crossing instead? 3). Similarly, the amplitude is just the raw data value at the peak, but what about e.g. peak-to-peak amplitude (you can kind of hack this right now by using the baseline-subtracted traces but even that's not quite right). 4). Currently spikes are detected over the whole trace - should be able to define a detection window.
May edit this with more things as I think of them