Measuring Length with TEM Images
- Open NIH Image.
- Go to the File menu, choose Import, select the TEM
image desired, and click on Open. This should bring up the TEM image and
a toolbar. Go to the toolbar and select the line selection tool. (It is the button fourth
from the top next to the paintbrush. It looks like a dotted line.)
- By clicking and dragging the mouse, draw a line along the length of the TEM image to be
measured. In the example below, the length of one of the gray spheres is being measured.

- Go to the menu bar, click on Analyze, choose Options,
and make sure the box next to Perimeter/Length is checked. If it is not
checked, click on the box to check it.
- Now go back to the menu bar and click on Analyze, but this time choose Measure.
This should bring up a window titled Info. Sometimes the Info
window is hidden behind other windows and so it is not visible. If this is the case, go to
the menu bar, click on Windows, and select Info. This
should make the Info window appear.
- In the Info window, look for the number next to the word Length.
This is the length in pixels of the line drawn in Step 3. Write down or remember this
number.
- This step will require a calculator. To convert the length obtained in Step 6 to
nanometers, use the following formula:

where length(nm) is the length of the line from step 3 in nanometers, length(pixels)
is the length of the same line in pixels, and cf is a conversion factor obtained
from the table below.
Conversion Factors
| Magnification |
Conversion Factor |
95% confidence |
| 500x |
34.42 |
0.06 |
| 660x |
26.72 |
0.08 |
| 870x |
20.51 |
0.04 |
| 1150x |
15.66 |
0.04 |
| 1500x |
11.72 |
0.02 |
| 1500xLM |
12.32 |
0.02 |
| 2000x |
8.99 |
0.03 |
| 2750x |
6.70 |
0.05 |
| 3800x |
4.82 |
0.03 |
| 5000x |
3.63 |
0.02 |
| 6600x |
2.85 |
0.02 |
| 8800x |
2.15 |
0.02 |
| 11500x |
1.58 |
0.03 |
| 15000x |
1.34 |
0.06 |
| 20000x |
1.00 |
0.03 |
| 27500x |
0.74 |
0.02 |
| 38000x |
0.55 |
0.02 |
| 50000x |
0.41 |
0.01 |
| 66000x |
0.31 |
0.01 |
Please note that the conversion factors are different for different
magnifications and that they vary linearly with the reciprocal of the
magnification.
The 95% confidence level is equal to twice the standard deviation between measurements
of the same image; 95% of all measurements made at a given magnification will deviate less
than this amount from the pixel value quoted.
Page 8 of 8 -