##CALCIUM This repo contains signal processing with calcium/BOLD signals
Evoked:
- Reverse raw data in time course(-1*raw data);
- Demean data, to eliminate 0Hz strong signal, and maybe a 10*log10(x) to be used to minimized dramatic signal drop;
- FFT timecourse signal into spectrum;
- Match to fmri by cut off signal before stimulation
- Extract targeted frequency band from spectrum, used as X-correlation input;
- Stimulation freq = 3.33Hz, and 1Hz for respiratory, 6Hz for cardiac; (harmonics exists)
Resting-State:
- Reverse raw data in time course;
- Demean data, and plot an pwelch spectrum to visualize the frequency-power profile.(10log10(x) applied);
- FFT timecourse into time-frequency decomposition(Power spectrogram);
- Extract frequency band with salient power;
- Apply an low BPF to power profile and see the correlation with BOLD;
##fMRI Normally, BOLD signal is observed via either AFNI(from NIH) or visualized on Matlab(Brucker packages). Now one more choice is to do so using Nipy(http://pysurfer.github.io/auto_examples/index.html)
From Nipy, signals can be ploted on a 3D reconstruction model of human brain.(Which indicates animal ones)