spO2 calculation

[dhans]

Hi, 

Can any body help me out of this dilemma? I need to know how to calculate spO2 from photo transistor output. In depth explanations are most welcome!!!

I have done this calculation before by using the below formula.

Ratio (R) = ((AC rms of Red LED)/DC of Red LED)) / ((AC rms of IR) / (DC of IR))

spO2 = 110 - 25xR

Is this correct?

[GerdW]

Hi dhans,

so you have some device named "photo resistor" without providing specific information (manual, spec sheet) for that device.

And you have a formula where you don't provide any information on (sources, datasheets, explanations).

You want us (in a LabVIEW programming forum) to discuss the validity of that formula for your (rather unknown) device without providing any in-detail information!? Really?

[dhans]

Hi buddy,

 Here I have given all details of my pulse oximeter project.

I am using nellcor DS -100 A probe which contains Red LED, IR LED and one photo transistor.

Using labview I alternatively on and off Red and IR LEDs. I save an output of photo transistor for both Red LED and IR LED in two different arrays. Then I implement an algorithm said above and some filtering process for smoothening the waveform. 

Since the output of photo transistor is current I needed to convert current to voltage using an op amp called ' MCP602' . This op amp has two channels. I took out output from 1st channel and connect it to the filter circuit in order to remove DC components from output signal. From that signal I do some calculations to get an expected output waveform. 

Just I need to check whether my algorithm and logic implemented in labview is correct and to know the proper alogrithm and chnages. 

[GerdW]

Hi dhans,

I need to check whether my algorithm and logic implemented in labview is correct and to know the proper alogrithm and chnages.

I cannot comment on your algorithm as I don't know anything about the specific physical properties you are using in your setup…

But I can comment on your VI:

- I don't see any error handling

- I do see a lot of unneded local variables

- I see some "funny" other things

I recommend:

- use a state machine approach like "INIT", "IR LED", "RED LED", "Analyse"…

- keep relevant values in shift registers (maybe in typedefined clusters) to get rid of local variables

- enforcing dataflow by outputting boolean values in case structures is just bs: that's why I recommend a state machine…

- implement error handling on your hardware communication functions!

[dhans]

Hi GerdW,

Thanks for the reply.