|
|
 |
  
 |
|
The longer the time between the exposure completion and reading-out becomes, the smaller the detection value is. Is this phenomenon what is called “fading”? |
 |
|
As the longer the time between the completion of exposure and the start of read-out, the lower the detected value becomes compared to when reading-out IP immediately after the completion of exposure. This phenomenon is called fading. |
|
|
Why does fading occur? |
|
|
In the IP that has been exposed to radiation source and such, electrons are excited and maintained in meta-stable state. In such a state, when electrons are stimulated by red light during the scanning process, they emit blue photostimulable luminescence. These photostimulable luminescence are detected electrically for a detected value. They, however, sometimes get thermal influence other than this detection process, through which some electrons in meta-stable state lose energy before scanning. This eventually causes the fading phenomenon. |
|
|
What quantitative characteristics does the fading have? |
|
|
Under the conditions of conventional autoradiography (exposure for more than 1 hour in room temperature), the detection value decreases at a certain pace despite the nuclide of the used RI. The higher the temperature is, the sharper the detected value decreases but the decrease of detected value cannot be detected at the temperature of 5 degrees or lower. The longer the exposure time is, the slower the detected value decreases. Although this is confirmed only with the BAS-III IP, since the fading progresses even during exposure, it is likely the case with other IPs for BAS currently sold by Fujifilm. |
|
|
What should we be careful about concerning the fading phenomenon? |
|
|
Since the autoradiography using IP generates fading phenomenon, we must be careful about the following points when IP is used for quantitative purpose. -There is no linearity in detected value of BAS system regardless of exposure time.(However, Dr. P. Fernyhough, Dr. B. R. Whitby reported at the Society for Whole Body Autoradiography Meeting (Michigan, USA, 1997) that linearity depending on exposure time can be obtained by using refrigerated shield box.) -When exposure and reading take place in different locations, and when comparing the experiment results from different experiments, it is necessary to arrange not only exposure times but the times between completion of exposure and scanning. -The extent of fading varies depending on temperature. Therefore, when the exposure time is less than 10 minutes, if the temperature of IP became higher than the room temperature while being in the IP eraser, it is better to cool it down to the room temperature by putting it into a cassette and such for stability of the data. -When the exposure time is as short as less than 10 minutes, the extent of fading varies very much depending on the time before the reading, which causes unevenness. To prevent this, reading must be started 10 to 30 minutes after the completion of exposure. |
|
|
Will there be IPs with which no fading occurs? |
|
|
In terms of technology and industry, it is unlikely that IP with no fading phenomenon will be launched in next 2-3 years. At this moment, it remains to be wise to avoid the influence of this phenomenon or to use some adjustment measures if necessary. |
|
|
The detected value does not change proportionately with time. |
|
|
The detected value of BAS system has no linearity against exposure time, which you must be careful about. No linearity specifically means that the detected value of 14C, or other sample with which the loss of RI can be neglected, after 10-hour exposure is not 10 times its detected value after 1-hour exposure (It will actually be 8 times). (However, Dr. P. Fernyhough and Dr. B. R. Whitby reported at the Society for Whole Body Autoradiography Meeting (Michigan, USA, 1997) that linearity depending on exposure time can be obtained by using refrigerated shield box.). |
|
|
Does a detected value have higher and lower limits? |
|
|
There are maximum and minimum values determined by the reading conditions of BAS system. For example, with sensitivity(S)=10,000 and latitude(L)=4, which is an usual setting, the maximum value is around 4,000 and minimum value (except for 0) is 0.4 when expressed in the PSL/mm2 dimension. The pixels higher than the maximum value are to be recorded as the maximum value and the pixels lower than the minimum value, as the minimum value. So, when the reading L=2, values expressed in the PSL/mm2 dimension, the maximum value is about 400 and the minimum value (except for 0) is about 4.0. When the exposed IP, that is supposed to show the PSL/mm2 in the range of 1.0 and 3.0, is read with S=10,000 and L=2, the image becomes 0 and is invisible. If the same IP is read under the conditions with S=10,000 and L=4, an appropriate image data can be obtained. |
|
|
How big is the A/D conversion error in a detected value? |
|
|
The error that occurs when converting continuous analog data into digital data that is discrete is called “quantum error.” BAS system keeps the quantum error at a certain ratio for the whole density range. When the reading conditionis latitude L=4 and Gradation=256, this error ratio is 3.7%, i.e., the original PSL of data with PSL/mm2=100 range from 100 to 103.7 and PSL/mm2=1.0 range from 1.0 to 1.037. With BAS system when the quantizing error becomes smaller, the number of gradation is larger and the latitude is smaller. With BAS-2000, therefore, the smallest error is about 0.2%. This value is between about 4.6 to 0.014% with BAS-2500 ( ) and BAS-5000. Since the image density detected as PSL/mm2 is addition of the background and the sample, the quantum error ratio of the image intensity (PSL/mm2-BG) becomes large when the background noise is high. For example, an image data read with BAS-2000 at latitude=4 and gradation=1024 has 0.9% of qunatizaion error. In case the detected value is PSL/mm2=100 and background noise is PSL/mm2=90, the sample density is (PSL-BG/mm2)=10. (It has been confirmed that even when the background contribution is 90% of the detected value, BAS has enough linearity for a 1x1cm sample with uniform concentration.) In such case, the quantum error is 0.9PSL/mm2. Therefore, when quantifing a very small amount of sample accurately it is effective to set the gradation as large as possible and lower the background noise using a lead shield box. Although lowering latitude is effective with regard to minimizing quantum error ratio, we do not recommend it for autoradiography applications because it narrows the quantitation range. |
|
|
How does a pixel size relate to resolution? |
|
|
A system such as BAS scans an object with laser beam to create a two dimensional image data. The pixel size is the pitch of the object surface (here, the surface of IP). It can be calculated by multiplying the sampling time and scanning speed. Therefore, the pixel size does not mean resolution. Resolution is affected by many factors such as machine, type of IP exposure, sample and so on. Resolution, however, will not be better than the pixel size. |
|
|
What causes an image lacking in resolution? |
|
|
Among BAS series, the smallest pixel size varies from 200x200µm to 25x25µm. Whichever BAS is used, its highest image resolution does not exceed that of X-ray film autoradiography. However, when BAS images are considered as “low resolution,” “rough image,” “lack of resolving power” the reason might not be of the BAS performance but of the radiation distribution from the sample. The number of disintegrated RI must be statistically treatable in order for the radiation amount to a pixel can be correlated to the RI distribution within a sample. In other words, the shorter the exposure time, the rougher an image becomes due to the statistical variation of the RI disintegration. BAS can generate images within 1/100 exposure time of X-ray film and the image obtained is quantifiable. Therefore, the exposure time in autoradiography using BAS tends to be short. When quantification result is only the matter of interest, exposure time can be short, however, for obtaining a high-quality image, exposure time must be long enough. |
|
|
How can you improve image quality? |
|
|
When resolution of the BAS image seems low, in most of the case this occurred because the exposure time was too short. The first thing you can do for quality improvement is extending exposure time. If extending exposure time does not make much difference, then it is effective to use machine with smaller pixel size. When using a machine that allows you to choose a pixel size, choosing larger pixel size will decrease the noise in the image and the resolution seems higher. Smoothing can be done by an image processing software such as L Process software for Macintosh and Windows. |
|
|
What should you be careful about when choosing a small pixel size? |
|
|
Choosing a smaller pixel size means that you need a high quality image. In order to obtain a high quality image with a small pixel size, the exposure time must be extended. Please extend the exposure time according to the reverse ratio of the pixel area ratio. |
|
|
Even after a long enough exposure, what if the image quality is not satisfactory? |
|
|
BAS series consists of machines with smallest pixel sizes of from 200x200µm to 25x25µm. Image quality might be improved by using smaller pixel size or using high-resolution type IP such as BAS SR IP or 3H IP |
|
|
