Scientific Volume Imaging b.v.
Iterative Constrained Tikhonov-Miller (ICTM) algorithm
For fluorescence and from a theoretical viewpoint Maximum Likelihood Estimation (MLE) Restoration Methods are the best choices. However, for high Signal To Noise Ratios (SNR) as in widefield images this becomes rather academic. The ICTM method in the Huygens Professional is computationally more efficient than the Classic MLE (CMLE) algorithm, but in most cases the Quick-MLE (QMLE) method is found to be still faster. Since QMLE is also good at handling high SNR widefield images, QMLE has superceded ICTM. For this reason the ICTM menu entry is now located under menu 'Restoration->Legacy->Iterative Tikhonov-Miller' in the Operations window.
As to the 'quality' of the result, that is either a subjective criterion based on visual inspection (ICTM might win in low noise cases) or it is based on a mathematically sound criterion in which case MLE is proven to be best choice. For noisy images like most confocal ones, MLE is not only scientifically more correct, it also produces visually more pleasing images: far less background noise artifacts than ICTM.
For further information, see this article.
See also...
<original> ictm <PSF> -> destination
?-it value (20)? ......... Max. iterations
?-sn value (30)? ......... Signal/Noise ratio
?-bg value (0.5)? ......... Estimated background
?-q value (0.1)? ......... Quality change threshold (%)
?-mode fast | highQ (fast)? ......... Iteration mode
?-pad auto | parent | padpar | fullpadpar (auto)? ......... Padding mode
?-centerPsf? ......... Center PSF
?-nv? ......... Silent
Returns: void.
Remarks:
The original image and the PSF image are both mandatory arguments. If the PSF image is blank, a theoretical PSF will be calculated automatically based on the provided microscopic parameters. See Point Spread Function on section →[x].
When using -pad parent no padding is actually done.
When using -centerPsf the center of mass of the PSF is aligned with the center of the image. By leaving it off small PSFs shifts due to differences in chromatic aberration between the channels in multi-channel images can be corrected manually.
No comments. Add one.
