Non-conducting specimens for SEM should be coated with conductive material, gold is one of the most common coatings. Looks like operator of your SEM coated specimens with too thick layer of gold and forgot to tell you about it. If possible, coat specimens with carbon. Do not pay attention to amounts of C, O and other light elements in your "quantitative" analysis, they are wrong.

@Vladimir Dusevich sir thank for your comment and advice.

Dear Vineet Kumar,

It does not seem very familiar to find gold in canine uroliths, much less in the concentrations that you indicate tens of per cent. Most likely, it is due to clear contamination during the preparation process or to an erroneous quantification employing M lines with severe interference. Good luck

It is very familiar to see gold in anything, because gold is a common coating material.

You haven't said whether you're carrying out your analysis in an electron microscope (in which case @Vladimir Dusevich is very likely correct) or whether you are using straight ED-XRF. If you can, check with a graphical display of XRF lines whether there is anything more plausible close to an Au line. I haven't got my display to hand, so I'm dependent on tables of data; the only half-way plausible overlap I can find is between Au Lalpha and Zn Kbeta, which are about 0.14 keV apart. Zn doesn't sound very plausible either, but I hope this gives you a clue as to how to proceed.

Check that your instrument is set up to search for biologically-plausible elements: the search software usually checks for only a limited range of elements based on the sort of sample it's expecting. I remember with much amusement the time when I'd given a moderately awkward sample to a company rep and he mis-identified Br as Al (Br Kalpha and Al Lalpha are very close), because his instrument was set up for the wrong type of sample.

Thank you @David Beveridge sir. In this analysis we have used scanning electron microscope.

In that case, have you any reason to doubt @Vladimir Dusevich's explanation?