Replies to post #324460 on Amarin Corp Plc (AMRN)

[rosemountbomber]

Thanks Lemm. I think your post is rather prescient because my wife is dealing with a number of things concurrently and needs to have a gum infection taken care of by an endodontist but now having seen a cardio because of this EKG, until the cardio does an echo and stress test he says that he can only categorize her as high risk even though like you he says to pay very little attention to that past EKG as he says it is not much value as a diagnostic, but now this may impede the endodontist from proceeding as the wife would need IV sedation. What a mess.

[sts66]

If it has no value then why does the algorithm even exist? I assume a doctor, not a nurse, looks at every EKG and they're perfectly capable of being able to read one without the computer telling them "look at this!"? Asked a different way, how would a doc analyze an EKG using his own brain and education and come to a completely different result than the computer's interpretation? In engineering things (thank god!) are pretty much black and white, there's not very many "maybes". Speaking of which, an analogy to this EKG algo is that the kinds of thermal systems I used to design and build were pretty much operated by hand on the ground (called joy-sticking) so engineers could observe things and see a problem developing in time to issue commands to prevent it from occurring - but once those systems are in orbit a computer must figure out what's going on and automatically takes steps to correct problems. So I helped the programmers write the control code for those systems, telling them "if you observe X, Y may occur, do Z to prevent it", and there were many scenarios that the algos had to be able to recognize and correct, took well over a year to get that work done. I assume a similar method was used to create that EKG machine algo to detect possible MI's - somebody(s) had to use the own personal observations and knowledge to write that algo, and higher ups should have reviewed and approved it - was it a case of garbage in garbage out, or are EKGs so hard to interpret they're nearly useless for anything but blatantly obvious problems? (meaning "maybes" are more frequent than yes or no)