john-judge / TSM-to-ZDA

Data acquisition with TurboSM and data processing with PhotoZ
MIT License
0 stars 0 forks source link

ZDA analysis reproducibility #1

Open john-judge opened 2 years ago

john-judge commented 2 years ago

https://github.com/john-judge/ImageZ/issues/2

john-judge commented 2 years ago

5-31-22: 5-trial averages

Overall, pixels of high SNR do not appear in groups. Occasionally in groups of 2-3 px.

1-0

image

1-0-14: 200 uA

Px Avg Max SNR 2.12 (binning = 1) 2.07 (binning = 3)

1-0-15: 0 uA

2.10 (binning = 1) 1.82 (binning = 3)

2-0

image

2-0-16: 0 uA

2.22 (binning = 1) 1.96 (binning = 3)

2-0-17: 200 uA

2.07 (binning = 1) 1.82 (binning = 3)

3-0

image

3-0-18: 0 uA

2.10 (binning = 1) 2.02 (binning = 3)

3-0-19: 200 uA

2.22 (binning = 1) 2.01 (binning = 3)

No significant max SNR difference over all recordings 2.14 at 200 uA, 2.14 at 0 uA (binning = 1) 2.00 at 200 uA, 1.97 at 0 uA (binning = 3)

Possibly an issue with pixel arrangement in TSM-to-ZDA load? Decrease measure window to width of 1, and show Max Amp, it should look like RLI frame? But with other sample ZDA taken from old rig, this isn't the case either.

john-judge commented 2 years ago

Fixed issue. Seems now unambiguously good result? Previous analysis to be repeated. image

Mid-stim SNRs, at 200 uA, of 2.45, 2.67, 2.61, compare to old rig of 1.43, 1.27, 1.33, 1.34, 1.44 (from 2/9/22 01_01_01, 01_01_02 and 5/12/22 01_01_01, 01_01_02, 02_01_01): image

The SNR_new / SNR_old ratio for these slices is thus 2.58 / 1.36 = 1.89

john-judge commented 2 years ago

image 01_01_17.zda