This is an interesting term to apply to digital signal processing. What (to me at least) suggests is that these different statistical approaches do little to change the data, but they impose personal “hopeful” results.  This is a direct quote from a very nice opinion paper on digital signal processing, but it just made me understand that the human brain, at this point, is still largely responsible for outcomes of those processing procedures.

“Statistics and probability are used in Digital Signal Processing to characterize signals and the processes that generate them. For example, a primary use of DSP is to reduce interference, noise, and other undesirable components in acquired data. These may be an inherent part of the signal being measured,…”

I am not saying that digital signal processing is not useful for helping to shape views about what signals (in this case, which peaks) are worth “recognizing as important” and those which are seen as “undesirable“, but the bottom line is still “it is ultimately my choice” and my application of the metrics used in that digital processing that ultimately “I accept”.

So I am not sure why so much credibility is given to these methods over the (yes laborious) task of “thinking” rather than just accepting what the apps put out.
This concept has been brewing for a long time, “the good the bad and the learned” which title i should change to “the good, the bad, and the undesirable” perhaps (LOL), and I have been trying to sort out how much “bias” i add to these apps as i work to find the number of peaks along a trimer (as height, width and valley data).

It seems that the question of the benefit of trying to be diligent at using algorithms for peak determination along a trimer of surfactant protein-D (at least at this point in time) is perhaps not that helpful (save a fairly robust verification that I see just a little more than what the peak finding apps see). I dont think this is heubris, I really think that at this point in time, the nuances of bilateral symmetry and huge differences in peak widths, heights, and peak positions along a plot are pretty much neglected (undesirable and overlooked) in the peak finding apps used so far  (scipy, octave -iPeak.m, AutoFindPeaksPlot.m, Lag-Threshold-Interference, PeakValleyDetectionTemplate.xlsx, PeakDetectionTemplate.xlsx, and others, using various settings (all these apps have been mentioned hundreds of times in this blog).

I am using those digital processing data, along with what i see as peak patterns as a blend for a final choice for peak parameters (yes that’s bias)(bias is a dirty word for ‘learned’).

A quote from the same article mentioned above “…… the final judge of quality is often a subjective human evaluation, rather than an objective criteria. These special characteristics have made image processing a distinct subgroup within DSP” (aside,  i think he should have used the singlar of criteria (criterion) here).