So if only iritis inhibits the scan, and that in only 5 of every 24 cases, and only 1 in 1000 have iritis, then that just means a whopping 5 in every 24,000 people would require manual screening to determine identification. I don't see the problem here.
IMHO, the problem is false positives.