GammaVision
®
v7 (A66-BW) 783620H / 1013
the key line has overlap with another library peak. If any of the qualifying key lines is not found, the key line test fails and the nuclide is rejected.
If the fraction limit is not zero (the default value is zero), the fraction of the branching ratios is calculated. If the calculated fraction is less than the Fraction Limit set on the Analysis tab, the nuclide is rejected. Branching ratios of all qualifying in-range peaks for a nuclide are summed.
Then sum of branching ratios for all qualified and identified peaks for a nuclide is summed as well. The ratio of the two sums is compared with the branching ratio limit. If it is less than the limit set, the nuclide is rejected.
If the “Fraction Limit Test flag” in b30winds.ini
is False and Directed Fit is enabled on the
System tab, the library peak flags are checked to determine if a peak is “qualified” for its branching ratio to be summed (in the total sum and the identified peak sum). Peaks with the “Not
In Average” flag set in the library are excluded along with peaks outside the analysis range.
6.3. Calculation Details for Peaks
For all library peaks in the analysis energy range, the program attempts to calculate the net peak area and centroid of a peak at that channel. At this step in the analysis, each peak is considered to be a singlet. A singlet is a single, isolated peak; that is, it is far enough away from other peaks in the spectrum so that the spectrum is background on both sides of the peak (does not overlap another peak).
6.3.1. Background Calculation Methods
You can select the method from among these types: automatic, 5-point average, 3-point average, and 1-point minimum (see Section 5.5.1.1).
6.3.1.1. Automatic
For the first pass, the peak centroid is the library energy (Fig. 233).
To calculate the first pass background on the low-energy side of the peak, the 5-point average of the channel contents is calculated for the region from the peak-centroid channel to the channel which is 6 times the library match width (normally 0.5) times the calculated FWHM (from the calibration) below the centroid. The 5-point average data at a given point is the sum of the data from two channels below the point to two channels above the point divided by 5. This is the same as smoothing the data with a smoothing width of 5 and coefficients of 0.2 for all points.
The background value is the minimum value of the moving 5-point average and the background channel number is the center channel of the 5. If the minimum average value is within one sigma
(counting statistics) of the actual channel value at the assigned channel point, this 5-point average is the low energy background value for this peak. If the average value is not within one
238
783620H / 1013 6. ANALYSIS METHODS
Fig. 233. Background Calculation.
sigma of the actual data, a 3-point average is used instead of the 5-point average to calculate a new minimum value. This 3-point average minimum value is compared with the actual data at the assigned channel and is accepted if it is within 1 sigma of the actual data. If the 3-point average also fails this test, the data value at the assigned channel is used for the background.
The same process is repeated for the high-energy side of the peak to calculate the background value above the peak. The background under the peak is the straight line between these two values.
The net peak area and background are calculated from this first pass. Next, the width is reduced or increased depending on the peak-area-to-background ratio and the library match width. This adjustment makes two improvements: (1) it reduces the number of channels in the peak for small peaks (decreasing the uncertainty), and (2) it improves the area calculation for peaks moved from the library energy.
This background calculation method (that is, automatically selecting 5-, 3-, or 1-point averaging, depending on which method best approximates the spectrum data) has advantages, when there are closely spaced peaks, over other methods. For example, because the 1-point method will be used when a small peak is very near a large peak, a more accurate measure of the background will be obtained as compared to the 5- or the 3-point average (Fig. 234).
239
