PIESNO

Probabilistic Identificatino and Estimation of Noise (PIESNO) is a technique developed for Magnetic Resonance Imaging (MRI). It is a simple technique for identifying noise-only pixels and estimating the noise level or parameter, which is also known as the Gaussian noise standard deviation. PIESNO is operationally very simple once the interested user has all the necessary numerical constants or parameters. The computation of these numerical constants is slightly more involved and I have created a MATLAB file as well as a GUI to facilitate interested users in obtaining the numerical value of these parameters. The relevant terms shown in the GUI are defined in the paper, Ref. [1].

MATLAB Interface

For those who want the ease of using Matlab functions and features, and want to use the PIESNO routines through Matlab, you are in luck. Thanks to Dr. Bonny and his request, I have created an MATLAB interface specifically for PIESNO, which uses the same sample data I used in our work, Ref. [1]. I assume that you already have hispeed.jar. Then, you can just download these files, data.bin and SampleTestforPIESNO.m. Please make appropriate changes in SampleTestforPIESNO.m about the directory paths where you keep hispeed.jar and data.bin. If you are able to run the program successfully, then you should get three figures as shown in the screen shot here.

Aside: NIH email server got so panaroid that it won't let me send "data.bin" through email. So, I put it here so that others can download it for testing PIESNO.

GUI

This GUI is for those want to create their own PIESNO algorithms.

To use this GUI, please download the jar file, and type the command "java -jar PIESNOParameters.jar". Or, you can just click here through Java Web Start.

A screen shot of this GUI (with Windows Look & Feel) can be gleaned here.

As you can see, there are three selections (inputs) and five outputs.

The order of the inputs and the outputs are:

(1) Input: Probability level (α) [ It is defined in Section 2.3 of Ref.[1] ]

(2) Input: The number of combined channel (N)

(3) Input: The number of images (K)

(4) Output: λ_– [ It is defined in Section 2.3 of Ref.[1] ]

(5) Output: λ₊ [ It is defined in Section 2.3 of Ref.[1] ]

(6) Output: The optimal quantile order (α^*) [ It should have been given a different Greek letter. Some of the values are listed in Table 1 of Ref.[1] ]

(7) Output: Denominator (based on the optimal quantile order) [ The values are listed in the last column of Table 1 of Ref.[1] ]

(8) Output: Denominator (based on the median) [ The values are listed in the last column of Table 2 of Ref.[1] ]

Finally, if you find the software useful and use the software in your research work, please spread the words by citation or word of mouth. Thanks.

References:

[1]. Koay CG, Özarslan E and Pierpaoli C. Probabilistic Identification and Estimation of Noise (PIESNO): A self-consistent approach and its applications in MRI. Journal of Magnetic Resonance 2009; 199: 94-103.