Hacker News Reader

Comment by eloisant | Hacker News Reader

I don't know, when you're talking about "the right arm", it seems obvious that it's the patient's right arm and not the arm that you see on your right when you're looking at the patient. I don't think you need to go to medical school to understand that.

View on HN

Assumptions like this is how errors are made.

Don't make assumptions, codify it.

You're right that "the left lung" should obviously be the lung on the patient's left. However, can you look at the image below and immediately identify which is the "left" lung?

https://scratch.rprosperoglobal.duckdns.org/lungs.png

I think I can, maybe. However I’m a radiographer and pictures like this are my job.

The orientation is screwed up and images should almost never be displayed like this.

If the patient’s organs are in non-standard locations (situs inversus, dextrocardia etc), something very bad will happen when orientations are messed with. Use standard orientations and label the images.

I fully agree with you. I was mostly posting as an example that "left is the patient's left" is not sufficient without other standards (like image orientation).

And that went straight over my head.

Your point was well made.

...you just expect random readers to know how lung looks on (i assume) MRI, from (I assume) top ?

I'd also assume it would be simple to spot just because the image is always took from the same direction so the narrow part points out to front/back of the patient

It’s obvious 99% of the time, but that 1% equates to a lot of fatal errors.

Reply on news.ycombinator.com