Chronic implants

[Julie-Fabre]

@emilyasterjones @cbimbo @pipcoen @cagjony @RikvDaal @mwawra @ajuavinett @thomaszhihaoluo @agbondy @mdmelin @jcouto @anemri

Hi all,

There is quite a bit of choice in terms of chronic implants for Neuropixels thanks to you all! I have gotten a few questions on the best implant for [x] use - a lot of users can find it a little tricky to navigate and understand the features of each implant and I am no expert myself. A useful thing for the community could be a table highlighting all the differences between the implants. @cbimbo has created this template of most implant's features / notable characteristics. Could you have a look and let us know if you want to add or change anything?

Feature\Lab	Churchland	Brody	Haesler	Ferraris	Carandini/Harris	Isogai	Aery-Jones	Couto	Buzsáki
Organisms	Mouse	Rat	Mouse/rat	Rat	Mouse/rat	Mouse/rat	Mouse	Mouse	Mouse/rat
Brain areas tested	Visual cortex, subiculum, midbrain	> 10 brain regions	Olfactory cortex	HPC, thalamus, amygdala	> 10 brain regions	HPC, MEC, V1, Amygdala	HPC, MEC	> 6 brain regions	HPC, ?
NPX version	1.0/2.0	1.0	1.0/2.0	1.0	1.0/2.0	1.0/2.0	1.0/2.0	2.0	1.0
# of probes	1	up to 4	up to 2 (in rats)	1	Up to 2 (in mice)	1	Up to 2	Up to 6	?
Implant weight	1.0 neuropixels: 1.2g2.0 neuropixels: 1.5g	2.6g each	2.1g	?	1.4g dual 2.0	2.6g	?	0.57g	0.87g
Weight w/ probes + estimated cement	2.0g	?	3.5g - 6.8g w/ HS	13g w/ HS	2.5g	3.5g 2.0 w/ HS	previous: <4g dual 2.0, new: <3g dual 2.0	?	?
Features	Lightweight, low-cost, optional cone implanted around implant	Custom positions, angles and depths for each probe	Lightweight, ease of use + assembly	Robust, high-retrieval success rate	Lightweight, low-cost, felxible, high retrieval success rate	Tested for social interactions	Custom positions, angles and depths for each probe	Custom positions, angles and depths for each probe	Drivable
Publicly available?	Yes - paper	Yes - paper	Yes - paper	Yes - paper	Yes - paper	Yes - paper	Yes - protocols.io	Yes - preprint	Yes - paper

PS: Please add the relevant people to this discussion! I used the respective github repos to get github handles of the people who have committed in the repos but I might have missed some people :) I'll edit this post as you all comment below. Edits: Added Emily's updated information, added Max's updated info

[agbondy]

Hi @Julie-Fabre. Yes, all that info is correct for our holder. Thanks for putting this together!

[emilyasterjones]

Thanks for organizing this; it's very useful! Two small corrections for my design:

1. My last name is Aery Jones
2. Implant weight is <3g (<4g was an older design, but I haven't updated the protocols.io to reflect this yet)

[mdmelin]

Thanks for putting this together! We also have a NP1.0 version (weighs 1.2g) and have implanted rats with a protective cone around the holder(s).

[cagjony]

Hi, Julie, thanks for reaching out,

As you know from our previous discussion, we have 2 different designs published in 2 different papers (van daal-nat prot and steinmetz-science). To overcome the confusion going around for several years, we can have 2 columns instead of one.

Our Nat. Prot. fixture was made only for np1 probes, and there are 3 types of fixtures

• 1 probe (freely moving mouse and rat with headstage included)
• 2 probes (rat only),
• drivable (rat only)

Our Science fixture was made for the np2 consortium (specially made for freely moving mice) and has the same design principles as the np1 fixture (smaller and lightweight). There is only one publicly available version, and it can carry 2 np2 probes (we also use it with 1 probe). The data from Figure 1E of the np2 paper comes from this fixture and from freely moving mice that carry 2 np2 probes with headstage included. (so rat only in that column is not correct - up to 2 with mice up to 4 with rats instead.)

All our design have tested more than 10 brain areas, they have high retrieval success rate and also have tested in social interactions (Eco-HAB).

About low-cost feature, we should either remove or add to all fixture designs in the table because all of them can be 3D printed and make them low-cost by definition.

I also think that users/builders/designers should state weaknesses/disadvantages of their system for example our np2 dual probe design is not flexible introducing steep angles and dual probe version has to be only implanted with having the same angle for both probes.

Best, Cagatay

[thomaszhihaoluo]

Wow, this table is really helpful. Thank you, Julie!

[anemri]

Hi Julie and Célian, thanks for putting this together!

For the Buzsaki implantation system: Number of brain areas tested > 10 (a few like HPC, V1 in the methods papers, the rest from over 80 labs using the system) NPX version: 1.0/2.0 Number of probes: 1 probe per drive. Up to 2 drives in mice, up to 3 in rats Implant weight: The implant alone weighs 0.47g (0.87g is if you 3D print the parts in steel, instead they are machined in aluminum) Features: Drivable, compatible with other chronic probes, high retrieval success rate

Best, Abdel