The following information differs from what appears in NeuroQuant's Triage Brain Atrophy Report with any structures that are highlighted in red. A brain structure is highlighted in red when atrophy for the structure is at the 5th percentile or lower, as compared to healthy controls of your same age and gender. So yes, those red areas are pointing out atrophy, but according to the following neuroimaging expert, atrophy occurs at larger volumes. I can't speak specifically about Neuroreader reports with respect to how they indicate atrophy, as I'm only vaguely familiar with Neuroreader. However, the percentiles for atrophy are the same for both technologies, per this neuroimaging expert.
The information is from Evolving Past Alzheimer's Episode #6: Is My Brain Shrinking? The interviewee is Cyrus Raji, MD, PhD, an expert in neuroimaging. He's an advisor for Neuroreader, but from what he has said just in this podcast, it sounds very much like NeuroQuant, at least in terms of how volumes are reported.
First, here's a paragraph from the summary in the show notes:
What do my MRI volumetric numbers mean?
To be considered abnormal and potentially consistent with Alzheimer’s diagnosis – the “magic number” for hippocampal volumes on MRI of the brain (using the Neuroreader or Neuroquant tool) is a number below the 25 percentile (this information is included in the report provided by Neuroreader or Neuroquant). Ventricular volumes, if higher than 75th percentile, may be too big and suggestive of Alzheimer’s or another disorder as well.
Note: I think the above summary should say "below the 26th percentile," given what Dr. Raji actually says in the interview.
If you're interested in the details, go to 30 min, 20 seconds in the podcast (of course, the preceding discussion is also interesting). But I transcribed Dr. Raji's words here. He says that the consensus is that:
for brain volumes, you should really be worried about volume loss and atrophy when a given brain structure like the hippocampus is only 25th percentile of normal or lower. So that means that if it's only 25% as large as we would expect it to be—or lower, like 20% or 15%—then that should be cause for concern of genuine volume loss in that area. And that number seems to be a good compromise between the sensitivity we want in detecting abnormality but also the specificity we want to make sure that it's a genuine abnormality and not some variation of normal. We don't want to make an over-call. So what happens if you're in the 26th percentile? Or 30th percentile? For me, I still raise the possibility that there could be something abnormal about that structure. I'll typically call that borderline-normal.
Typically, most of the brain structures we see are in the median; they're 50th percentile. So if you see that number, you shouldn't be concerned and say, 'oh my goodness, I'm only 50% of the norm' because that's still a very healthy number. And sometimes you see people that are really brilliant, it seems, and they have 99th percentile hippocampal volumes, and they're very large. You can see that as well, and that's not abnormal. It's just abnormal if it's too low.
Then for the ventricular volumes, those fluid-filled spaces that exist in the brains for housing cerebral spinal fluid, we typically call those abnormally enlarged if they're higher than 75th percentile of normal. So when brain structures shrink, that's not good. When ventricular structures get bigger, that's not good either. They're both indicative of the underlying atrophy process. So we use the 75th percentile for the ventricular volumes and the 25th percentile or lower for the other brain structures.
He then goes into asymmetry, which I have in multiple brain structures... I need to listen again to understand the significance of this phenomenon.