[MagnumPIG]

This is not really "huge", sorry.

For one thing, it's still an intrusive method which requires brain surgery. I expected something like radioactive isotopes but no, this is easily summed up as "better electrodes". As good thing for sure, not a game changer.

For another, the writer is a tech journalist and therefore more than likely wrong or inaccurate on important facts.

[hexxiiiz]

I am not aware of any work that has gotten a good timeseries of neuromodulator response. So much speculation is thrown around about how these hormones respond to real world experiences, but the evidence has been coarse grain from things like PET scans or sparse assays. Even if the tech used is not that new, the kind of data is. I am unaware of any data like this. Maybe I missed something. Is there another way neuroscience has managed to get an actual neuromodulator signal?

[skohan]

I would not under-count this advancement.

One thing which struck me back when I was studying neuroscience was just how impactful new tools were. If you look at the history of neuroscience, major leaps forward have almost always been driven by a new tool: whether it's a new dye, or a new imaging technology, every time we have gotten a better look at how the physiology of the brain actually works, it has elucidated topics we could only speculate about before.

So it's easy to poo-poo something like this from the sidelines, but if we can get more detailed information about the behavior or neurotransmitter concentrations over time, even if it's just a point measurement this could be an input to more accurate models which could help explain all kinds of things about how the brain actually works.

[wombatmobile]

Yes, not huge. Just a point measurement, achieved through invasive surgery and electrical stimulation.

What we really need is a schematic diagram. It’s not obvious how we will get to there from here. Not knowing anything about the data format means these are still early days.

But every little bit tells us something more, and one day iall the bits will all add up.

[hexxiiiz]

We actually have a bit of a schematic diagram already. It is well known that both dopamine and serotonin are largely produced by clusters of neurons in specific fore/mid-brain regions that then project out into the rest of the brain. The axions of these modulator neurons secrete the hormone, and it is known for both serotonin and dopamine (1) what receptors they bind to, (2) what transporters reuptake them into the presynaptic neurons, (3) what enzymes break them down. So the schematic of the system can be sketched out, and details of it, such as receptor functioning, can be determined pretty well in vitro. What we don't have is the timeseries data to properly parameterize this model. This research starts to provide this. You should check out research in pharmacology; there is a big hole for this and people are just guessing based on this schematic what the dynamics are.

[vanderZwan]

> What we really need is a schematic diagram. It’s not obvious how we will get to there from here.

The rise of gene-based identification of cell types, allowing for much more fine-grained understanding of how the cells are connected and develop than morphological identification alone, is a pretty good start I'd say.