Computerized gamma-spectrometry.

Many of us old timers can remember our first visit to the Chalk River Nuclear Laboratories and viewing a 100-channel kick-sorter. It was quite an impressive sight, filling a large room with 100 individual scalers with their neon lights flashing away.

One of us (MDS) as a graduate student used a 256-channel instrument. Compared to the older kicksorter, it was highly miniaturized, fitting into a pair of 19-inch panels six feet high. This great leap forward resulted from the use of computer technology to make one device do the work of 256.

On the negative side, the colorful term kicksorter dropped into oblivion to be replaced by the more functional multi-channel analyzer or MCA. The beast was a highly temperamental device that devoured premium-grade vacuum tubes and spent more time down than up. It was replaced by a transistorized model roughly the size of two PCs stacked upon each other. Its superior reliability enabled the completion of a PhD thesis.

About a decade later while setting up the laboratory at Ontario Hydro's Bruce nuclear generating station "A", MDS purchased another [Gamma]-spectrometer. This one filled only 3 to 4 feet of a single 19-inch instrument rack and had 4,096 channels.

The large increase in the number of channels was required as a result of the amazing improvement in resolution that came with the changeover from Tl-activated NaI scintillation detectors to Li-drifted Ge crystals. This instrument also had an integrated DEC PDP-11 computer which could analyze a full 2,048-channel spectrum in about 20 minutes.

For this month's review, we decided it was time to have another look at what had transpired over the years in [Gamma]-spectrometry. [Gamma]-spectrometry is a monitoring technique vital to fields such as nuclear medicine, health physics and nuclear plant operation. In some ways, [Gamma]-spectrometry has changed little. The principles are the same. The equipment does the same things with better reliability, linearity, pulse pile-up rejection, etc. and the ability to perform the analysis automatically.

The starting point for this review was a visit to Aptec Engineering. APTEC is a fully integrated Canadian manufacturing, sales and service company that builds complete systems from detector to data analysis. The 4,096-channel machine mentioned above had been purchased from them and they had given good service.

In addition, we had the opportunity to read an IAEA (International Atomic Energy Agency) report that surveyed eight [Gamma]-spectrometry software packages. The testing took identical spectral data and applied several criteria to rank the packages with respect to how well qualitative and quantitative analyses were performed. APTEC's software came out the best in the overall ranking.

Following a tour of the facilities by Aptec's vice-president, sales, Phil Stark, we were provided with a [Gamma]-spectrometer to take home as a souvenir . . . well not quite a complete one. There are some obvious problems walking out the door with radioactive sources and a liquid-[N.sub.2] cooled detector.

What we did have was the software to run a 16,384-channel system complete with some sample spectra on a single disk. The complete hardware to replace the old 256 channel analyzer (or more precisely to replace 64 of them) fits onto one plug-in board for a PC.

The package of brochures and manuals was comprehensive and included a single pink page that told us how to get "a running start" without reading any of it. That title comes from Aptec's president Ed Zieba who is a runner and knows that there are two types of people who operate software: those that read manuals and follow them to the letter and those who will never read a manual and would rather fumble their way on their own. He has successfully looked after both.

Using only that pink page, the software was installed and with a click-click here and a click-click there, we were ready to analyze a spectrum of fallout from Chernobyl. Pressing "P" for peak search, started it going. Then pressing "M" did a multiplet fitting of overlapping peaks, "I" identified the nuclides and "Q" quantified the identified species. In a few seconds, we had a full quantitative analysis.

Now, it was time to play around and see what the system could do. Using that pink page along with an occasional peek into the manual, it was possible to become fairly proficient in a very short period.

A simple click of the mouse in the left corner where it said "lin", changed the display from linear to logarithmic. A second click changed it to square root. A menu system allowed adding a grid or changing the color.

We produced some spectacular displays, but had to control ourselves when the spectrum was to be printed by a black-only printer.

The x-axis could be set for channel number or energy. As the cursor was moved along the spectrum, the numerical values for both the channel number and energy at that position were displayed above it. For each press of the spacebar, the cursor advanced from one ROI (region of interest) to the next, landing on the centroid of the peak.

The identity of the nuclide was displayed in a panel below along with the quantity present and a collection of statistical data such as minimal detectable activity, standard deviations and peak FWHM (full width half maximum).

A second panel showed acquisition data such as time, date, counting time and dead time. It was a simple matter to change the units for the quantitative measurements. User definable choices included activities in dps, Ci, Bq or multiples thereof ([Mu]Ci, MBq) per mL, kg, etc.

The ability to split the display into two or more windows was a very useful feature. One window could display the full spectrum while the other looked at a smaller band in a much expanded mode. This made it possible to view the detailed structure of each individual peak to check for overlapping peaks or some anomalous shape. Incidentaly, the IAEA report ranked the APTEC software as being the best for resolving multiplets.

The ability of a [Gamma]-spectrometry package to function is very dependent upon the quality of the energy-efficiency calibration. APTEC allows for various calibration curves up to fifth-order polynomials, fifth-order log-log or cubic splines. The calibration curve could be seen on-screen or in a graphical or tabular format to check the goodness of fit. It was quite impressive to see that all the points from the standards fit dead on the curve. There were no outliers or near misses. In addition to the calibration, this software also provides higher-order fitting for background subtraction.

Aptec are selling systems worldwide. Before we were allowed to leave the facilities, Rob Goble insisted that we have a tour of his health physics area. We saw an assembly line for hand-and-foot monitors. Today, these machines talk to you rather than wait for you to read the instructions. If your hand is in the wrong place it will tell you. The one we tried spoke to us in Korean. Not only did it do the monitoring job, the on-board computer maintained records of its operation.

Would we recommend the Aptec [Gamma]-spectrometry software? Ontario Hydro answered that one for us. MDS had an opportunity to visit his former laboratory in January. The old 4,096-channel MCA had been retired due to old age. In its place sat a pair of new APTEC systems. During the tour of Aptec, we had the opportunity to see a gaseous fission product (GFP) monitor being tested prior to shipment. Ontario Hydro didn't buy one; they bought several.

PCMCA/WIN OSQ/Plus, version 5.3: Aptec Engineering Ltd., East-50B Caldari Road, Concord, ON, L4K 4N8; Tel: [phone]; Fax: [phone].

The names of hardware and software mentioned in this review are trademarks of their respective developers.