A Basic Positron Emission Tomography System Constructed to Locate a Radioactive Source in a Bi-dimensional Space

The overall goal of the following experiment is to show the basic principles of a positron emission tomography system using two pairs of photo detectors. The main advantage of using this prototype over commercial PET systems is that users can see clearly the basic working principles of PET. We first had the idea for this system when we sought to elaborate our lab activity for master students, in the medical physics field.

Demonstrating the procedure will be Marcus Fontaine, a technician from my laboratory. Begin work on the PET setup by preparing the detection system. Make each detector using a photo multiplier tube and a polished plastic scintillator piece.

Design the scintillator piece to match the size and shape of the photo-multiplier tubelight entrance to allow good coupling. The next step is to wrap the scintillator in black tape. Leave one side uncovered to be coupled with the photo-multiplier tubelight entrance.

Put the scintillator aside and begin working with the photo-multiplier tube. Use alcohol and a clean cloth to clean the photomultiplier tubelight entrance. When this is done, get optical grease and apply it to the light entrance until it is covered completely.

Return to work with the scintillator and apply optical grease to its uncovered surface. After the optical grease has been applied align the photo-multiplier light entrance and the uncovered scintillator surface. Then couple the two treated surfaces.

To complete the detector, wrap the photo-multiplier tube and the scintillator in more black tape as illustrated by this example. Create at least four detectors before moving on to test them. Test the detector by connecting it to a voltage source and to a standard digital oscilloscope.

Turn the light off and on in the lab while monitoring the signal on the oscilloscope to verify there are no light losses. Next, connect the photomultiplier tube output to a scaler to count detected events. Adjust the detector voltage to its lowest value;in this case point five volts.

Now turn attention to the detector. Place the sodium 22 source directly on the surface of the scintillation portion. Record the number of detected events over three minutes using the scaler.

Then increase the control voltage by point zero one volts. Repeat the time to measurement of detected events for this value and all others up to the maximum control voltage. After all detectors have been completely tested, make a plot of the number of detected events versus the control voltage.

This plot includes data from 12 detectors. Select two detectors with similar curves to form a coincidence pair. For example, detectors O and I in this plot.

After two matched pairs are identified, move on to build and test two coincident systems. Construct the coincident system by placing two matched detectors so the scintillator part of one detector is directly over the scitillator of the other. Next, check the electronic components.

The coincidence system requires three nuclear instrumentation modules, a discriminator, a logic unit, and a scaler. Connect the output signals of the detectors to the inputs of the discriminator module. On the logic unit, select the end mode on the front panel.

Then connect the two discriminator outputs to the logic unit inputs. Connect the logic unit output to the scaler module to count the events. In this configuration, the scaler should register coincidence events due to cosmic rays hitting the coincidence system.

This schematic is of the completed coincidence detection system for testing a pair of detectors. Test each of the pair of detectors before proceeding. The next step is to create the PET setup.

Place the four detectors at the corners of a square. Orient the scitillators to point toward the center of the square. The detectors that directly face each other form a coincidence pair.

The detector surfaces are about 20 centimeters apart. Continue by connecting each of the detectors to separate channels of a discriminator, a logic block, and a scaler. For coincidence counting with four detectors, implement the logic block depicted in this diagram.

Here the detectors are labelled I, J, E, and F and their signals are passed through a discriminator that is not shown. In this diagram, each detector is associated with an OR module. Connect the output of each detector to the inputs of the three OR modules to which it is not associated as indicated by the colored lines.

Next connect the OR module's outputs to an AND module to complete the coincidence logic. In the following steps, supply power to only one coincidence pair. Then repeat the steps for the second pair.

Return to the detector array with a sodium 22 radiation source. Place the source along the line between the detectors of the coincidence pair. Use one detector as a reference and measure the distance to the source.

At the electronics rack, use the discriminator outputs for the two detectors as inputs to an oscilloscope. Display the signals on the oscilloscope with time on the horizontal access. In addition, take the output of the coincidence logic block and input it into the oscilloscope.

The resquare signals should appear on the screen. Use time differences between discriminator signals measured on the oscilloscope for later calibration steps. Next, prepare the computer automated measurement and time to digital converter or TDC module.

Connect the output of the logic block to the TDC start input. Then connect the outputs of the discriminator to a delay module. Pass the output of the delay module to the TDC's stop inputs.

This schematic provides an overview of the setup at this point. Connect the modules and computers with a general purpose instrumentation bus for software control. Install Labview or similar software to create a virtual instrument.

At this point, perform TDC calibration steps for the coincidence pair. The next step is to calibrate the system beginning with one coincidence system. Along the line between the two detectors, identify five equally spaced points including the midpoint.

This setup has points at four centimeter intervals. Next, place the radiation source between the points. With the source in place, begin to collect coincidence events, TDC data for 30 minutes.

Plot the average of all the collected TDC data every one or two minutes. This is a typical plot after the full 30 minutes. The two curves correspond to the two detectors of the coincidence pair.

Use a plot of the average of the difference of the TDC values from the two detectors to define a time zone for the purpose of locating the source. Now, move the radiation source to a different marked position between the detectors. Acquire 30 minutes of data and determine a new time zone and repeat for each of the five points.

Next, begin to test the sensitivity of the pair. Repeat TDC measurements at all marked points for different collection times. After testing sensitivity move on to set up a virtual instrument.

The instrument will associate a software LED with a position of the radiation source. The five vertical LEDs in the center of this array correspond to the five marked positions used to define the time zones. In this figure, our TDC data for two adjacent positions.

The mean of the TDC difference data for a position is at the center of an interval. The interval represents the standard deviation of the data. The data should be from the collection time that will be used to resolve the positions two minutes in this protocol.

Choose a window of times measurement to associate with a position with an LED by truncating the intervals so that they do not overlap. Setup will be complete after both coincidence pairs are calibrated. At this point, there should be nine marked positions between the detectors.

Place the radiation source on one of the positions. Collect data for two minutes and test the accuracy of the virtual instrument in locating the position of the radiation source. These are plots of the response for each detector and a pair that form a coincidence system.

The cumulative average TDC value over a one minute interval is given on the vertical axis. The index of the one minute interval is along the horizontal axis. Data was collected over 30 minutes.

The different colored lines correspond to different positions of the radiation source. The difference between these average values is plotted here. Linear behavior is expected for each set of measurements allowing conversion between time and distance.

Note that there is increased stability as the acquisition time increases. These plots provide a measurement of the sensitivity of coincidence systems. The median and the mean of the TDC counts are plotted along the vertical axis.

The index of the position of the radiation source between the detectors E and F is plotted along the horizontal axis. The data used for each point are collected for five minutes. The position index one corresponds to the radiation source being approximately centered.

As the index increases, the source moves toward the F-detector. This coincidence system exhibits good spatial resolution since the curve for the median and mean diverge as a function of position. By comparison, the median and mean for this system does not diverge as quickly indicating poor spatial resolution.

Once completed, this device can display the location of a radiation source in one minute or less. While attempting this procedure, it is important to remember to put the radiation source in a position where one pair of detectors is sensitive. Don't forget that working with radioactive material can be extremely hazardous.

Train in radiation safety to learn the precautions to take while performing this experiment.