I have two series of medical images each one from different imaging modalities.

According to that, I have been segmented the Region of interest (the object which appears in both modalities )using U-net in these two series of medical images.

Now I want to measure the similarity between these two series of an image depending on segmented the Region of interest

1- How can I do this? or what is the suggested method or technique?

2- If these two series of the image spatially misalignment, How to measure the similarity between them?

  • $\begingroup$ What type of series do you have? What image modalities are involved? $\endgroup$
    – Aelius
    Mar 24 at 14:21

1 Answer 1


It is hard to answer without you giving specifics (e.g., what are the modalities, what is the segmented ROI). But if you have two images (e.g. CT and MRI, or T1 MRI and T2 MRI), and you segment some organ on both of them (e.g. lung or liver or brain) you could simply compare these segmentations. For that you could first align two images (if they are from the same patient rigid registration will do the job), and then simply compute Dice coefficient between the masks.

You can not meaningfully compare intensities of a ROI of different imaging modalities, this just do not have any sense, bc, MRI for example measures magnetic resonance, CT measures tissue attenuation, etc., these are completely different signals. But you could compare shapes of the objects the way I described above.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.