GammaVision
®
v7 (A66-BW) 783620H / 1013
6.2.2.1. Guidelines for Selecting an Analysis Engine
1) Decide which peak search method is best for your application.
2) Establish the sensitivity for the peak search (Peak Cutoff for library directed; Sensitivity for
Mariscotti).
3) Decide if an accurate analysis of the multiplets is important. If so, select NPP32 or the
ENV32 analysis engine.
4) Notice if there are false negatives in your answers. That is nuclides have been initially detected and later dropped. If so, revert to the WAN32 analysis engine.
6.2.3. Library Reduction Based on Nuclide Rejection (ENV32 and GAM32 Analysis
Engines Only)
At the start of an analysis using the ENV32 or GAM32 engine, a raw peak search is performed on the spectrum. The list of peaks found is compared to the peaks in the library, and the nuclides not likely to be present are removed from the library. After this library reduction step, a librarydirected peak search is performed similar to NPP32 (for ENV32) or WAN32 (for GAM32) to find and fit peaks, then quantify nuclide activity. After this analysis, the raw data (Mariscotti) peak search is run again to pick up any peaks not accounted for in the library-directed search.
Finally, if Directed Fit is turned on, any peaks in the library that were previously rejected are now fitted using Directed Fit formulation (Section 6.3.2.2) and the activity is calculated for nuclides that were previously not reported. If the “ENV factor” in b30winds.ini
(page 448) is zero, this test is not performed.
6.2.3.1. Details of Library Reduction
Following is a more detailed description of the library reduction phase of the ENV32 and
GAM32 analysis:
1) If the first peak for a nuclide in the library is not found meeting the peak cutoff, then that nuclide is removed from the analysis.
2) If the first peak for a nuclide in the library is found meeting the peak cutoff limit then each subsequent peak for that nuclide (to a maximum of three peaks per nuclide) is evaluated to determine if the nuclide should be removed from the analysis as follows:
3) For each subsequent peak i (to the maximum of three per nuclide), evaluate as follows:
236
783620H / 1013
Go to the next peak,
Else, a) Calculate the expected peak area for the i th
peak as:
6. ANALYSIS METHODS
b) Calculate the critical level L
c
for the i th
peak as:
If A
i
> L
c
and the i th
peak is not a valid peak (that is, the peak is not found or the peak uncertainty is greater than the peak cutoff), then the nuclide is removed from the library.
Next peak i.
where:
E
1
Br
1
E i
Br i
A
1
A i
= efficiency for the first valid peak
= branching ratio for the first valid peak
= efficiency for subsequent library peaks
= branching ratio for subsequent library peaks
= activity calculated for the first nuclide peak
= activity calculated for subsequent library peaks
B i
= peak background for subsequent library peaks
W
= peak width for subsequent library peaks
M
= number of total points outside the peak used to calculate background
ENV = user-adjustable ENV factor from the
Files\GammaVision b30winds.ini
file located in
C:\Program
(see page 448 in Section A.2.2). The default setting is 0.2, but it can be set higher to keep more nuclides in the library by minimizing the number of subsequent peaks evaluated using the formulation above.
6.2.3.2. Details of Library Reduction Based on Key Line and Fraction Limit Test
Like the ENV test discussed above, this library reduction test is performed at the start of an analysis after a raw peak searched is performed. In this library reduction algorithm, both the key line and fraction limit tests are carried out for ENV32 and GAM32 engines.
For the key line test, the first in-range peak for a nuclide is always considered as a key line, even if it is not marked as a key line. All key lines for a nuclide must be found in the spectrum, even if
237
