Microradiography

###Simulation of the microradiography evaluation with the microCT: Determine lesion depth and deltaZ.

Introduction

Quantitative microradiography is a valuable tool in the evaluation of de- and remineralization processes of enamel and dentin. A well known system is the TMR system from Inspektor NL. This technique has it merits. However, the availability of microcomputed tomography devices makes it very attractive to use the microCT data to determine the de- and remineralisation of dental tissues repeatedly and non-destructive.

The 3D evaluation is very attractive. Sometimes,however, the simple evaluation according to the same rules as the TMR would be helpful, too. There is no commericial tool available for this task. Therefore I invested some time and developed a plugin for the open source image analysis program ImageJ.

Usually I prefer to use Fiji, which comes with most plugins, which is need, in one package. The plugin is beta. But it should already work fine.

To use my plugin, download it and copy it to the folder „plugins“ in the Fiji or ImageJ directory.

You can use the program for free. I put it under theGPL licence for the moment.

When you use the program, please cite the plugin:

KHKs MicroCT_goes_MicroRadiography developed by Karl-Heinz Kunzelmann, Dept. Oper Dent, LMU Munich, Ger.

If you like it invite me to an Espresso or a Coke Zero when we meet at a scientific meeting!

The source code is part of the distribution. It would be very kind, if you improve the program, to send me your improved version so that I have some benefit, too.

Author

Karl-Heinz Kunzelmann

History

13.1.2011 public release

Source

Contained in KHKs_microCT.jar

Installation

Copy KHKs_microCT.jar into the ImageJ > plugins directory and restart ImageJ.

MicroCT_goes_MicroRadiography

This is a short description, how to use this plugin.

To use the plugin you have to select the slice which you want to evaluate.

Mark a line perpendicular to the surface which was de/remineralized, start at a certain distance from the surface to get background data, extend the line into sound tissue to have enoughsample points for the reference plane.

Fig. 1: Draw a line in the grey value image
Fig. 1: Draw a line in the grey value image

Execute the plugin:

Plugins > KHKs MicroCT Tools >MicroCT_goes_MicroRadiography_3

Fig. 2
Fig. 2

The necessary instructions are summarized at the bottom of the window. Along the vertical axis you can see the values of the gray value, the x-axis represents the pixels along the selected line. The curve is the gray value profile along the line which you selected.

Calculation of the microradiographically equivalent values is easy:

Mark the reference area = the sound tissue with the mouse clicks 1 and 2. Use the cross cursor and select the points so that approximatly the same area between the horizontal line of the cross cursor and the line is above and below the line.

Fig. 3
Fig. 3

After two mouse clicks it looks like this:

Fig. 4
Fig. 4

The black horizontal line is the average gray value of all points within the intervall you selected with the mouse. The upper red line is 5 % below the black line (the value of the black line is set to 100 %). For the evaluation select the first intersection between the blue line representing the gray values and the red line when you approach the lesion from the sound dentin. The lower red line should be a help for you to identify the transition between the lesion and the background. The lower red line should be equivalent to the 20 % level of the black line. For the calculations of the area always the gray value is used which is the closest to your mouse clicks.

Fig. 5
Fig. 5

You get the summary information in the lower part of the chart. But you can also copy the data to the clip board and paste then into any program (like OpenOffice Spreadsheet, Excel, SPSS etc.). To copy the value to the clipboard: File > Show Table >Copy to Clipboard. If you need the variable names for your first row in the spread sheet use: File > Show Table > Copy Header to Clipboard.

It is a good idea to add additional information about the treatment conditions and treatment group into this table.

Fig. 6
Fig. 6

Some remarks

ImageJ offers the possibility to calibrate the images. If you decide to calibrate the images, then the column „lesionDepth“ will be in the unit you selected (i.e. µm). If you do not calibrate ImageJ then the unit is „pixel“. To obtain a metric information you will have to multiply „lesionDepth“ with „pixelWidth“.

The value of deltaZ is the area which is marked in gray.

It is calculated by subtracting at each pixel in the gray interval the value of the actual gray value from the 95 % value of the average gray value (= avg * 0.95 - y-value at the location i).Then each difference is multiplied with the pixel width (for calibrated images is could be µm, for uncalibrated images it is 1pixel). Then all areas (difference times pixel width) within the gray intervall are summarized (= sum[i]).

This is the correct unit of the program.

The real unit depends on the unit of your gray value. It can be uncorrected 16 bit values or calibrated values for the linear attenuation coefficient. Our Scanco microCT offers the following options:

If the scan was calibrated for bone, then the voxel value is in units of Hydroxyapatite density [mg HA/ccm], otherwise in linear attenuationcoefficient [1/cm], Hounsfield [HU] or native file number [1] units.

This program is open to any units. You just have to know what you do and decide yourself.

Let me comment on the long number after the comma (the decimal places). Programming is a necessity and a hobby for me. It is not meaningful to use the whole number which was calculated by the program. I would round it in the program, but you can also easily round the numbers yourself in the spreadsheet or statistics program. It would need some time of focused work for me - which nowadays is nearly impossible at universities - therefore I left the long number. But I know - and you know it now, too - that only rounded values are meaningful.

With these explanations you should be able to interpret the unit of avgMineralLoss yourself !