2017-04-19--Notes_for_new_program.txt

EAG: einzelnes EAG
Settings are almost the same as GcEad

need to trigger for a signal response

t: set the total duration, maybe 10s
pre-trigger time: 1-2s, typically, during which signal is recorded prior to trigger

Measured maximum deviation in mV, duration of deviation, and the steigung into and out of the deviation (less important for most)

a project has many experiments, and an experiment has many recordings.
EAG should be associated with a stimulus

Later, we compare the responses

Others do this:
one recording that takes 60s, and multiple stimuli are given during this time.
The software looks for the maximum deviation after the trigger.
Trigger is recorded via one of the digital inputs.
Then the deviations are split into multiple recordings.
So one experiment would be one long recording with many experiments.

When a trigger is initiated, the signal gets zeroed out for a new baseline.

Recording number get assigned automatically.
Recordings get IDs, which is editable.  This labels the stimulus.
Stimulus description can also be set by the user.
The ID is used for analysis later.

Normalizing:
The antenna response degrades over time. E.g. might respond with 10mV at the beginning, then only 5mV afterwards
So you normalize by fitting a regression to the "aging" and then normalize the measurements over time.
This is done using a control stimulus (which is a "blank" paper that the antenna still responds to).

Also, if you reuse a stimulus, its concentration gets lower over time.  So you need to do the same normalization
using a reference stimulus that causes a strong response.

Then you can check the other stimulus and what would change if you used that as the reference.

---
User interface

There are three windows.
If you click on the project, you see a table of StimID, Stim Description, Count (how often did you record that stimulus), average (mean of amplitudes), standard deviation, minimum, maximum
The project also has a histogram view (the reference stim may be highlighted in a different color).

From the experiment, you see a thumbnail of each recording.
Three views: thumbs, table, historgram
Table: stimId, Stim description, recording (max amplitude), surface(v*s) (area of the deviation, but we don't need it), maximum, threshold (not needed), slope1 (v/s), slope2
Histogram: for just this experiment

Clikc on recording, and you see the recording osciloscope view.
A vertical line for the trigger.
A horizontal line for the zero where the trigger was.
A dot fot the min
A dot for the middle of each slope, and a line for each slope.
The dots need to be movable, and then recalculate automatically.

In the experiment view, you can deselect or delete bad recordings.

Remarks can be entered in the project and experiment views.

In EAG you need to test a lot and a lot of replications.
So you need to import a simulus list of the compounds you work with.

In the current EAG screen for recording osciloscope, the window is too technical.  Should be simpler.
Should be more like GcEad.  It should be easier to adjust the view of the signal (the zoom and offset)


antenna, stimulus, order

an antenna can be possibly used for multiple experiments if it lasts for a long time,
so track time of experiments as well.

Uncontrollable factors: age of antenna (so try to use insects with same age),
physiological status, mated or not.

Add: calculation of required experiment size in order to get desired CI.

Feature: shape recognition for each ID, to have the signals overlayed, to see if there are dramatic variations

Feature: graph of amplitude vs stimulus (ANOVA visualization)
Feature: graph visualizing whether different stimuli are significantly different than each other
Featuer: either suggest the randomization order of stimuli, or let the user specify it
Feature: would be nice to be able to change colors and line thicknesses on graphs
Feature: work with all the IDACs and the new one
Feature: export functionality (raw data, analysis data, graphs, tables)

PET123
ELL049