Data processing: measuring – calibrating – or testing – Measurement system – Statistical measurement
Reexamination Certificate
2000-07-12
2001-12-04
Hoff, Marc S. (Department: 2857)
Data processing: measuring, calibrating, or testing
Measurement system
Statistical measurement
C702S104000, C356S326000
Reexamination Certificate
active
06327549
ABSTRACT:
STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH OR DEVELOPMENT
Not Applicable.
BACKGROUND OF THE INVENTION
1. Field of Invention
The invention relates to a system for accurately determining the true counting rate of energy peaks in a radiation spectrum measured by a spectrometer. More specifically, the system extrapolates the live count based upon the length of the inactive period of the spectrometer to create a histogram representative of the radiation spectrum measured by the spectrometer.
2. Description of the Related Art
Spectroscopy systems are used to obtain the spectrum of a radiation field. The radiation might be charged particles, X-rays, or gamma rays. The invention disclosed herein is discussed in terms of a gamma ray spectrometer but applies to other spectrometers as well. The spectrum is a plot of measured counts as a function of the energy of the gamma ray as determined by the detector. Monoenergetic gamma rays produce peaks in the spectrum. The centroid of the peak denotes the energy of the gamma ray. The number of counts in the peak, suitably corrected for the detector efficiency, the counting geometry, and the counting losses in the spectrometer, gives the counting rate, or activity, of the gamma ray.
The spectrum is collected by measuring the energy of gamma ray photons one at a time. After amplification and filtering, the amplitude of the detector pulse is converted to a digital value. The measured value, proportional to the photon energy, is used to point to a location in a histogram memory array. The memory location, called a channel, is used to store the counts for a narrow range of energies. The digitized amplitude points to the appropriate channel and the contents of that channel are incremented for each count.
All spectrometers have some effective dead time per pulse that causes the measured number of counts to be less than the true number observed by the detector. Reducing the spectrometer dead time can reduce the errors caused by counting losses but generally makes the energy measurement less precise. Usually the dead time is adjusted to a value that gives adequate energy resolution and some method of correcting for the counting losses is applied.
The simplest correction scheme uses a true timer and a live timer. The true timer counts a precision oscillator during the differential data acquisition. The live timer counts the same oscillator but is gated off while each pulse is processed. The area of the peak divided by the live time is the calculated count rate. This method can give accurate results if the time during which the live timer is gated off accurately reflects the dead time of the spectrometer. This method is well known in the art, as disclosed by R. L. Chase,
Nuclear Pulse Spectrometry
, McGraw-Hill, New York, 1961.
One very accurate technique for implementing a live timer is the Gedcke-Hale method. Jenkins, Gould, and Gedcke describe the Gedcke-Hale method in Chapter 4 of the text by Ron Jenkins, R. W. Gould, Dale Gedcke,
Quantitative X-ray Spectrometry
, Marcel Dekker, New York, 198 1. Gedcke noted that, statistically, the dead time per pulse is twice the time-to-peak of the pulse plus the time required to return to baseline. In order to doubly weight the time-to-peak of the pulse, the Gedcke-Hale method counts the live timer backward during the rising part of the pulse then gates it off during the falling time.
Live timer techniques, such as Gedcke-Hale, give accurate results if the counting rate throughout the spectrum is constant during the acquisition. Some measurements, particularly in the field of neutron activation analysis, involve isotopes with half-lives shorter than the measurement interval. In this case, the count rates in some peaks are very high at the beginning of the measurement and then reduce as the isotope decays. Since the live timer measures the live time for the entire measurement interval, it can not correctly account for the large number of counts lost early in the measurement.
J. Harms,
Nucl. Instr. And Methods,
53:192 (1967) describes a method for accurately determining the true count rate in spectra involving peaks from isotopes of differing half-lives. The central idea of the Harms method involves making an estimate of the ratio of the actual arrival rate of pulses to the measured rate, hereinafter referred to as the true-to-live ratio, r. When a photon is recorded, the appropriate channel is incremented by r. For example, if half of the counts are being lost during dead time intervals, then r is two. When the digitized value points to a memory channel, the channel contents are incremented by two instead of one. In effect, the lost count is accounted for by accumulating the measured count twice. Although this procedure might appear to be incorrect since the lost count did not necessarily have the same energy as the measured count, Masters and East, IEEE Trans. Nucl. Sci., 17, 383, showed that in typical spectra involving many thousands of counts, the resulting error is statistically small. Harms noted that r could be obtained from the spectrometer's true time and live time clocks.
Westphal (U.S. Pat. No. 4,476,384) improved on the Harms method by introducing the concept of a virtual pulse generator. The virtual pulses do not actually exist in the spectrum, which would cause errors, but are subject to the same dead times as real pulses. The pulses are counted in two counters, one continuous and one gated off during dead times. The gating off period is equal to the pulse processing time extended by a fixed interval equal to the pulse rise time. The ratio of the two counters gives an estimate of r and is used to increment memory just as in the Harms method.
Hereinafter, a method that corrects for counting losses by adding an increment other than one to memory is referred to as a differential correction method (DCM). The main advantage of a DCM is that it can correct for varying count rates in different peaks. A major problem with a DCM is the loss of statistical information. It is well known that counts in a non-corrected spectrum follow normal statistics. If the number of counts in a peak is C, the standard deviation of the counts is the square root of C. In a spectrum obtained using a DCM, the error is unknown and can not be deduced from the spectrum alone.
Accordingly, there is a need for a spectroscopy system which can both accurately determine the count rate for spectra containing peaks with differing decay times and maintain the necessary information about the spectra to allow the statistical error to be calculated.
It is the object of the invention to produce a spectroscopy system which can give correct counting rates for spectra containing peaks with differing decay times without losing statistical error information.
BRIEF SUMMARY OF THE INVENTION
It is not presently known how to estimate the error in a peak count C using only the information in the DCM corrected spectrum. However, in the present invention, two spectra, the DCM spectrum and an error spectrum, are simultaneously accumulated. The two spectra have the same size, true acquisition time, and energy calibration. The substantive difference between the spectra is in number of counts in each channel. In the DCM spectrum the selected channel is incremented by r when a count of that energy is measured. In the error spectrum the selected channel is incremented by r
2
when a count of that energy is measured. In the data analysis the best estimate of the number of counts in a given peak comes from the number of counts in the DCM corrected spectrum. The statistical error is obtained from the error spectrum, possibly corrected by other known factors obtained during the acquisition. The use of r
2
to obtain the error spectrum is justified in the detailed description of the invention.
REFERENCES:
patent: 4476384 (1984-10-01), Westphal
patent: 4661918 (1987-04-01), Buschmann et al.
patent: 5050065 (1991-09-01), Dartois et al.
patent: 5349193 (1994-09-01), Mott et al.
patent: 5608222 (1997-03-01), Hardy, II
patent: 5884234 (1999-03-01), Jorion
Bingham Russell D.
Gedcke Dale A.
Keyser Ronald M.
Trammell Rex C.
Twomey Timothy R.
Barbee Manuel L.
Hoff Marc S.
PerkinElmer Inc.
Pitts & Brittian P.C.
LandOfFree
Differential correction method and apparatus does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Differential correction method and apparatus, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Differential correction method and apparatus will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2575580