Hi..

Please go through the following links:

https://www.med.emory.edu/EMAC/curriculum/diagnosis/sensand.htm

https://onlinecourses.science.psu.edu/stat507/node/71

Thanks,

Sobhan

This is a good paper explains the first part of your question:

https://umanitoba.ca/faculties/health_sciences/medicine/units/pediatrics/sections/neonatology/media/DiagnosticTest1.pdf

Hi..

Please go through the following links:

https://www.med.emory.edu/EMAC/curriculum/diagnosis/sensand.htm

https://onlinecourses.science.psu.edu/stat507/node/71

Thanks,

Sobhan

Thank you professor, I am working on bio-signal steganography .can you please help me with the second question. how to find sensitivity and specificity for a stegnography technique.

I guess you mean steganalysis, Correct?

In steganalysis we have two classes, positive corresponding to stego samples, and negative corresponding to cover samples.Now we define the following measurements (taken directly from this article : Comprehensive Review of Audio Steganalysis Methods, IET Signal processing, 2018, DOI: 10.1049/iet-spr.2016.0651):

- True negative (TN): the number of cover samples that are classified as cover samples.

- True positive (TP): the number of stego samples that are classified as stego samples.

- False negative (FN): the number of stego samples that are classified as cover samples.

False positive (FP): the number of cover samples that are classified as stego samples.

True positive rate (sensitivity) is the probability of detection of stego samples correctly and it is calculated as:

Sen = TP/(TP+FN)

True negative rate (specificity) is the probability of detection of cover samples correctly and it is calculated as:

Spe= TN/(TN+FP)

Accuracy (Ac.) is the probability of correct classification and it is calculated as:

Ac = (TP+TN)/(TP+FN+TN+FP)

for the second question you need to generate a database using your steganography method, then implement a relevant steganalysis method and apply it on your database.