new drdt episode spoiler. culprit discussion
guys i think it just clicked in my brain how ace being the culprit would also make a lot of sense and i don't like the idea of him dying BUT since teruko will be onto him in the next episode and i'm pretty condident there's still a lot of the trial left then surely SURELY ace's name will be cleared right. he WILL survive this trial i swear to g
30 notes
·
View notes
hello! i would love to hear the rant about PET scans :3
Holy shit so okay I'm in the train for the next 20 minutes, and I _know_ that's not enough time to get into all of it, but I could rant about this for hours, so. Maybe we cap this at 20 minutes. [20 minutes later] Okay so I wrote a huge wall of very boring text that barely started getting into it, so let me provide way less detail, actually:
It is massively expensive. A PET scanner costs a lot. And it's not a one time purchase, and then you can do scans, no, you wish. You also need some very expensive equipment to create radioactive tracers (which are what is used to do a PET scan) on site, because that stuff needs to be created fresh (under an hour) before every scan. To create the tracers is ALSO incredibly expensive. A single PET scan costs multiple thousands.
This also means that PET research makes use of as few participants as possible. A study with 15 participants is considered big. You simply cannot infer from 15 participants to the whole population. This also means that, statistically, it is highly likely that you don't find an effect even though it exists - meaning if your PET study looks for the effect of A on B, it is highly likely that it will find that A has no effect on B even though it does - simply because you didn't have enough participants (if this explanation doesn't make sense, let me know, and I can explain in detail)
This, together means, that an absolutely absurd amount of money is used for research that, by design, will not find results, because to find results, they would need more participants and even more money.
Because scientific publishing is a shitshow at the moment, research that doesn't find results very rarely gets published, especially not if you can't even be sure whether the result is right. So absurd amounts of money put into research that doesn't even get published.
And I haven't even talked about the results they did find and issues with them. Don't ask me to explain those. Don't tempt me to put hours into writing a multiple page essay that nobody will read.
So, in conclusion: PET is an absolutely amazing feat of engineering that is magnificent in detecting cancer and with it we could learn so. Much. More about the brain and how it works. But to do that, a lot of the basic organisation of how we do science would first need to change. Many labs would have to collaborate and be okay with making the collected data openly available, so appropriate sample sizes (=numbers of participants in a study) can even be achieved (Here's a paper on that). That probably won't happen, though.
Now, obligatory note: one of the professors who taught me about PET is a man who wrote an extremely controversial paper about exactly this stuff, despite also using PET in his research. If you like niche drama in science, look into this paper and all the articles that are responding to it.
9 notes
·
View notes