The detect duplicates feature in Rayyan is one of the best I have seen. However, never presume it does a perfect job of catching all the duplicates. The only way to ensure that is to manually go through your articles.
To start go to your review.
1) Click on detect duplicates, two blue boxes will populate indicating that the detect duplicate search has begun and how many duplicates were found.
2) Click on 'unresolved'. The unresolved duplicates will now populate as the list of articles. Click on the first article.
3) All possible versions of the article which Rayyan suspects as being duplicate will populate. Here we can compare two such suspected articles. The title, authors, journal are all the same and it is safe to say that this is a duplicate. Keep the top article, and click on the 'delete' icon for the second.
4) In the 'possible duplicates' box, located on the top left hand side of your screen, you can now see that there are only 76 articles left unresolved and that 1 has been removed.
5) Nothing is ever truly deleted on Rayyan, which is a nice failsafe, the article is simply removed from the main review and placed in a duplicate folder.