Baseline and peak area calculation

Share icon

Wed, 02/03/2016 - 08:27 am

Hi all,

I just want to ask/point two things.....1). I have an envelope with two overlapping peaks and i want to define the peak area under each individual peak. I have some a priori knowledge about positions and FWHM so i can apply some constraints. I also apply a linear baseline using two points on the left and right of my envelope and then i do curve fitting including the baseline in my fitting. As a result, Igor gives away all the parameter values including peak areas. I guess these peak areas are related to the actual area of each peak considering the baseline so it corresponds exactly to my actual Raman signal.....is that right? (The baseline i include because there is some background involved probably due to florescence). Do other people here do this the same way?

2). I have been playing around constraining the a and b values of my baseline function.....However, i have a difficulty to understand sometimes how Igor is considering the a value which if i am not mistaken is the intercept of the line.....What is the reference point (starting point) that takes into account when you fix various a values like -5 or 5 for example? I hope this is not too complicated.


Thanks a lot
Fitting a sum of two peaks plus a line for the baseline is a pretty common way to do this.

The a value of a line fit is the Y intercept, that is, the Y value when x=0. If your data X values are far from zero, the intercept can be pretty confusing:)

John Weeks
WaveMetrics, Inc.
support@wavemetrics.com

February 3, 2016 at 05:27 pm - Permalink

Hi John,

Thanks very much for your response......The reason why i asked about the peak areas is because i accidentally saw another post of another guy (Mr. Fosi) back in 2009 stated as ''Integrating area under curves'' where he wanted to do a similar thing....There was one reply from a guy basically describing what i described above but when you replied you mentioned something about needing some modelling.....Maybe i am missing something and the guy wanted to do something a bit different but this is what confused me and i wanted to make the post above.

February 4, 2016 at 01:19 am - Permalink

Mr_Fosi's post talked about using the Integrate operation to analyze his peaks. I suggested an approach that uses curve fitting.

Have you discovered the Multipeak Fit 2 package? It does all the fitting you describe, and has a relatively nice GUI.

John Weeks
WaveMetrics, Inc.
support@wavemetrics.com

February 4, 2016 at 10:10 am - Permalink

Hi John,

Well yes.....I use the multipeak fit 2, i add my expected peaks using some constraints (e.g. width or position) that i should also expect.....Then i do the fitting using a linear baseline and igor gives me the results....It contains all the parameters i need including the peak areas.....obviously i use the baseline because i have some background which is due to fluorescence...as far as i understand the parameter values are given taking into account the baseline which is fitted simultaneously....I prefer this approach rather than removing the background and then perform fitting to the corrected spectrum. I guess this is what you meant as well using multipeak fit 2.....am i right?

February 5, 2016 at 05:40 am - Permalink