video
Test your ability to differentiate real and AI x-rays

Alex Hogan/STAT
A recent study tested whether or not a small group of radiologists could tell the difference between AI generated X-rays and real ones. The results were interesting, albeit a little worrisome — doctors correctly differentiated the real from deepfake images about three-quarters of the time. STAT's Katie Palmer, who wrote about the study last week took the same quiz and scored about the same.
In the latest STATus Report video, Katie tells Alex how she could differentiate the images as well as the trained professionals. She says it has to do with two factors: noise and consistency within the images. Watch the video now to take the test yourself alongside Alex, who may have scored better than everyone.
addiction
Testing an antidepressant for meth use
As Lev Facher's War on Recovery series outlined, lifesaving medications for opioid addiction already exist — though access to those medications is extremely limited. For people with methamphetamine use disorder, there are no such drugs. But a study published yesterday in JAMA Psychiatry showed how an existing drug may be able to offer some relief.
The study found that compared to a placebo group, people who took an oral antidepressant called mirtazapine saw a greater reduction in how often they used meth. Among nearly 350 participants who had used meth almost daily in the previous month, those who took the antidepressant used an average of 7 fewer days per month at the end of a 12-week period. Those who got the placebo used 4.8 days fewer per month.
While the study size and the drug effect are both small, the researchers found the numbers encouraging. Still, it's worth noting that 23% of participants stopped taking mirtazapine due to adverse reactions like drowsiness or weight gain, compared to 15% in the placebo group.
No comments:
Post a Comment