Probe Geometric and hc-11 eeg

[gmbcrazy]

Hi, I'm trying to get into hc-11 data sets, some of the data were recorded by BZ64-6x10 probe. I'm quite confused how I could match the eeg data to the Probe Channels.

file='Achilles_10252013\Achilles_10252013.eeg'; lfp = bz_LoadBinary(file,'frequency',1250,'start',0,'duration',10,'nChannels',134);

Thus, lfp is a Time*134 matrix. Each row is a channel. How the row in the matrix corresponds to the probe channel (NeuroNexus)?

When I load the .eeg data in neuroscope, for that file, Chan 60-63 (Neuroscope Chan) extra channels in the probe, corresponds to NeuroNexus Channel 36-39 (Should it be Intan Chan in the NRX_Buzsaki64_6X10.xlsx?). However, in NRX_Buzsaki64_6X10.xlsx, it corresponds to Ch48-51.

[brendonw1]

Does this help: http://neuroscope.sourceforge.net/UserManual/data-files.html#data-file

The .eeg file falls under the "data file" type on that page

On Tue, Apr 17, 2018 at 5:49 PM, Lu Zhang notifications@github.com wrote:

Hi, I'm trying to get into hc-11 data sets, some of the data were recorded by BZ64-6x10 probe. I'm quite confused how I could match the eeg data to the Probe Channels.

file='Achilles_10252013\Achilles_10252013.eeg'; lfp = bz_LoadBinary(file,'frequency',1250,'start',0,'duration',10, 'nChannels',134);

Thus, lfp is a Time*134 matrix. Each row is a channel. How the row in the matrix corresponds to the probe channel (NeuroNexus)?

When I load the .eeg data in neuroscope, for that file, Chan 60-63 (Neuroscope Chan) extra channels in the probe, corresponds to NeuroNexus Channel 36-39 (Should it be Intan Chan in the NRX_Buzsaki64_6X10.xlsx?). However, in NRX_Buzsaki64_6X10.xlsx, it corresponds to Ch48-51.

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/buzsakilab/buzcode/issues/182, or mute the thread https://github.com/notifications/unsubscribe-auth/ADXrTUuX0Yq3-OfLpxKthI7Ky_asiIXGks5tpmOAgaJpZM4TZGk6 .

[brendonw1]

I just wrote the entire email in italics below below and then I remembered something, I think Andres Grosmark actually rearranges all channels so 1 is the highest on shank 1 and 2 is the second highest on shank 1 etc. So it might already be in order of

1. Shank 1 Superficial site . . .
2. Shank 1 Deepest site
3. Shank 1 Superficial site . . .
4. Shank 1 Deepest site
5. Shank 1 Superficial site . . .
6. Shank 1 Deepest site

I don't know what they did with the extra 4 channels, maybe they made them another group at the end of each probe??

If that scheme doesn't work then read below

Oh neuronexus... hm... yes those might be right, it depends on which probes were used. The files you cite, such as NRX_Buzsaki64_6X10.xlsx are relatively new and correspond with intan recording systems. While I did not record this data, it may be that that file corresponds with something like this:

https://docs.google.com/spreadsheets/d/1iTZo6ELsLXoYVg_j8EMgcrLbzFxMcepAb2BME4R4T50/edit?usp=sharing

If that isn't right, you'll have to figure it out based on this information:

This Buzsaki64Sp here should show you the combination of probe shank/channel layout and also how each channel pins out to the output pins: https://drive.google.com/file/d/0B0WCzFWl7GCXdVJMajhDc0RsODhjMFBTTkk2dHA4dlNFTmJz/view?usp=sharing

Then pages 25-26 show the preamplifier the probes were plugged into... the numbers here specify the numbers in the amplifed recording : https://drive.google.com/file/d/1yIc84Cssu1HExGEfOzx-f52UcupRfeD9X01SoAdC5fd0SLIJyJFhT_I5hlR1/view?usp=sharing

On Thu, Apr 19, 2018 at 10:21 PM, Brendon Watson brendon.watson@gmail.com wrote:

Does this help: http://neuroscope.sourceforge.net/UserManual/data-files.html#data-file

The .eeg file falls under the "data file" type on that page

On Tue, Apr 17, 2018 at 5:49 PM, Lu Zhang notifications@github.com wrote:

Hi, I'm trying to get into hc-11 data sets, some of the data were recorded by BZ64-6x10 probe. I'm quite confused how I could match the eeg data to the Probe Channels.

file='Achilles_10252013\Achilles_10252013.eeg'; lfp = bz_LoadBinary(file,'frequency',1250,'start',0,'duration',10, 'nChannels',134);

Thus, lfp is a Time*134 matrix. Each row is a channel. How the row in the matrix corresponds to the probe channel (NeuroNexus)?

When I load the .eeg data in neuroscope, for that file, Chan 60-63 (Neuroscope Chan) extra channels in the probe, corresponds to NeuroNexus Channel 36-39 (Should it be Intan Chan in the NRX_Buzsaki64_6X10.xlsx?). However, in NRX_Buzsaki64_6X10.xlsx, it corresponds to Ch48-51.

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/buzsakilab/buzcode/issues/182, or mute the thread https://github.com/notifications/unsubscribe-auth/ADXrTUuX0Yq3-OfLpxKthI7Ky_asiIXGks5tpmOAgaJpZM4TZGk6 .

[brendonw1]

Yes, I don't know who's asking but if it is in reference to the novelty sessions on CRCNS repo the channel numbers have been rearranged to reflect their physical layout (please read through the 'Dataset Description' and 'Channel_Orders' pdf's associated with the dataset - I attached them for reference). Hope this answers your question.

Best, Andres

On Thu, Apr 19, 2018 at 10:32 PM, Brendon Watson brendon.watson@gmail.com wrote:

I just wrote the entire email in italics below below and then I remembered something, I think Andres Grosmark actually rearranges all channels so 1 is the highest on shank 1 and 2 is the second highest on shank 1 etc. So it might already be in order of

1. Shank 1 Superficial site . . .
2. Shank 1 Deepest site
3. Shank 1 Superficial site . . .
4. Shank 1 Deepest site
5. Shank 1 Superficial site . . .
6. Shank 1 Deepest site

I don't know what they did with the extra 4 channels, maybe they made them another group at the end of each probe??

If that scheme doesn't work then read below

Oh neuronexus... hm... yes those might be right, it depends on which probes were used. The files you cite, such as NRX_Buzsaki64_6X10.xlsx are relatively new and correspond with intan recording systems. While I did not record this data, it may be that that file corresponds with something like this:

https://docs.google.com/spreadsheets/d/1iTZo6ELsLXoYVg_ j8EMgcrLbzFxMcepAb2BME4R4T50/edit?usp=sharing

If that isn't right, you'll have to figure it out based on this information:

This Buzsaki64Sp here should show you the combination of probe shank/channel layout and also how each channel pins out to the output pins: https://drive.google.com/file/d/0B0WCzFWl7GCXdVJMajhDc0RsODhjM FBTTkk2dHA4dlNFTmJz/view?usp=sharing

Then pages 25-26 show the preamplifier the probes were plugged into... the numbers here specify the numbers in the amplifed recording : https://drive.google.com/file/d/1yIc84Cssu1HExGEfOzx- f52UcupRfeD9X01SoAdC5fd0SLIJyJFhT_I5hlR1/view?usp=sharing

On Thu, Apr 19, 2018 at 10:21 PM, Brendon Watson <brendon.watson@gmail.com

wrote:

Does this help: http://neuroscope.sourceforge.net/UserManual/data-files.html#data-file

The .eeg file falls under the "data file" type on that page

On Tue, Apr 17, 2018 at 5:49 PM, Lu Zhang notifications@github.com wrote:

Hi, I'm trying to get into hc-11 data sets, some of the data were recorded by BZ64-6x10 probe. I'm quite confused how I could match the eeg data to the Probe Channels.

file='Achilles_10252013\Achilles_10252013.eeg'; lfp = bz_LoadBinary(file,'frequency',1250,'start',0,'duration',10, 'nChannels',134);

Thus, lfp is a Time*134 matrix. Each row is a channel. How the row in the matrix corresponds to the probe channel (NeuroNexus)?

When I load the .eeg data in neuroscope, for that file, Chan 60-63 (Neuroscope Chan) extra channels in the probe, corresponds to NeuroNexus Channel 36-39 (Should it be Intan Chan in the NRX_Buzsaki64_6X10.xlsx?). However, in NRX_Buzsaki64_6X10.xlsx, it corresponds to Ch48-51.

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/buzsakilab/buzcode/issues/182, or mute the thread https://github.com/notifications/unsubscribe-auth/ADXrTUuX0Yq3-OfLpxKthI7Ky_asiIXGks5tpmOAgaJpZM4TZGk6 .

[brendonw1]

Sorry, I meant to send this pdf (attached) instead of the 'Recording Summary'. -AG

On Thu, Apr 19, 2018 at 11:32 PM, Andres Grosmark andres9713@gmail.com wrote:

Yes, I don't know who's asking but if it is in reference to the novelty sessions on CRCNS repo the channel numbers have been rearranged to reflect their physical layout (please read through the 'Dataset Description' and 'Channel_Orders' pdf's associated with the dataset - I attached them for reference). Hope this answers your question.

Best, Andres

On Thu, Apr 19, 2018 at 10:32 PM, Brendon Watson <brendon.watson@gmail.com

wrote:

I just wrote the entire email in italics below below and then I remembered something, I think Andres Grosmark actually rearranges all channels so 1 is the highest on shank 1 and 2 is the second highest on shank 1 etc. So it might already be in order of

1. Shank 1 Superficial site . . .
2. Shank 1 Deepest site
3. Shank 1 Superficial site . . .
4. Shank 1 Deepest site
5. Shank 1 Superficial site . . .
6. Shank 1 Deepest site

I don't know what they did with the extra 4 channels, maybe they made them another group at the end of each probe??

If that scheme doesn't work then read below

Oh neuronexus... hm... yes those might be right, it depends on which probes were used. The files you cite, such as NRX_Buzsaki64_6X10.xlsx are relatively new and correspond with intan recording systems. While I did not record this data, it may be that that file corresponds with something like this:

https://docs.google.com/spreadsheets/d/1iTZo6ELsLXoYVg_j8EMg crLbzFxMcepAb2BME4R4T50/edit?usp=sharing

If that isn't right, you'll have to figure it out based on this information:

This Buzsaki64Sp here should show you the combination of probe shank/channel layout and also how each channel pins out to the output pins: https://drive.google.com/file/d/0B0WCzFWl7GCXdVJMajhDc0RsODh jMFBTTkk2dHA4dlNFTmJz/view?usp=sharing

Then pages 25-26 show the preamplifier the probes were plugged into... the numbers here specify the numbers in the amplifed recording : https://drive.google.com/file/d/1yIc84Cssu1HExGEfOzx-f52Ucup RfeD9X01SoAdC5fd0SLIJyJFhT_I5hlR1/view?usp=sharing

On Thu, Apr 19, 2018 at 10:21 PM, Brendon Watson < brendon.watson@gmail.com> wrote:

Does this help: http://neuroscope.sourceforge.net/UserManual/data-files.html#data-file

The .eeg file falls under the "data file" type on that page

On Tue, Apr 17, 2018 at 5:49 PM, Lu Zhang notifications@github.com wrote:

Hi, I'm trying to get into hc-11 data sets, some of the data were recorded by BZ64-6x10 probe. I'm quite confused how I could match the eeg data to the Probe Channels.

file='Achilles_10252013\Achilles_10252013.eeg'; lfp = bz_LoadBinary(file,'frequency',1250,'start',0,'duration',10, 'nChannels',134);

Thus, lfp is a Time*134 matrix. Each row is a channel. How the row in the matrix corresponds to the probe channel (NeuroNexus)?

When I load the .eeg data in neuroscope, for that file, Chan 60-63 (Neuroscope Chan) extra channels in the probe, corresponds to NeuroNexus Channel 36-39 (Should it be Intan Chan in the NRX_Buzsaki64_6X10.xlsx?). However, in NRX_Buzsaki64_6X10.xlsx, it corresponds to Ch48-51.

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/buzsakilab/buzcode/issues/182, or mute the thread https://github.com/notifications/unsubscribe-auth/ADXrTUuX0Yq3-OfLpxKthI7Ky_asiIXGks5tpmOAgaJpZM4TZGk6 .

[gmbcrazy]

Thank you so much!! @brendonw1

I also e-mailed Peter yesterday, I should use Amplipex instead of Intan mapping. Together with your information, I start to figure out the probable correct way.

I believe the channel is organized from superficial to deep in the .eeg file as you mentioned. Attached .jpg shows the LFP traces in one Shank (Animal Buddy, 8x8 probe), ripple power (blue font, RMS calculated from whole session) increase as Channel ID increase. If I'm correct, that suggest the deep channel is also in center of pyr layer. Or-else, I wonder if the ripple power could be even stronger when recorded in deep pyr layer instead of center layer?

BTW, the google sheet you shared should be correct. The one in buzcode-master\preprocessing\probeGeometries\NRXtoAmplipex_Buzsaki64_6x10_untested.xlsx might be wrong. Start from K23.

[brendonw1]

This is not my recording or dataset so I think I probably can't help more. If you want to email me I can help you find the authors who can give more details. But yes, if ripple-band power is greater you're closer to pyr layer. That said, I'm not sure I see that here in the blue traces... are the red ones ripple-band filtered and the blue are raw? In general you should also see units near the PYR layer and you should see a change in the sharpwave from slightly upgoing above the pyramidal layer to clearly downgoing below it. Here I see the blue trace showing upgoing half-waves during ripples in the top channels, indicating they are probably above the PYR layer and by the bottom the ripples are riding on flat to downgoing half-waves, implying you are at or beneath the PYR layer. So you may be right.

On Fri, Apr 20, 2018 at 10:47 AM, Lu Zhang notifications@github.com wrote:

Thank you so much!! @brendonw1 https://github.com/brendonw1

I also e-mailed Peter yesterday, I should use Amplipex instead of Intan mapping. Together with your information, I start to figure out the probable correct way.

I believe the channel is organized from superficial to deep in the .eeg file as you mentioned. Attached .jpg shows the LFP traces in one Shank (Animal Buddy, 8x8 probe), ripple power (blue font, RMS calculated from whole session) increase as Channel ID increase. If I'm correct, that suggest the deep channel is also in center of pyr layer. Or-else, I wonder if the ripple power could be even stronger when recorded in deep pyr layer instead of center layer?

[image: rippleexample] https://user-images.githubusercontent.com/32208252/39057195-a4766ed0-4486-11e8-874c-0f0c935d7200.jpg

BTW, the google sheet you shared should be correct. The one in buzcode-master\preprocessing\probeGeometries\NRXtoAmplipex_Buzsaki64_6x10_untested.xlsx might be wrong. Start from K23.

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/buzsakilab/buzcode/issues/182#issuecomment-383120389, or mute the thread https://github.com/notifications/unsubscribe-auth/ADXrTYXsdl-6F8AKoR1_Hwwlc8lHOxllks5tqfUZgaJpZM4TZGk6 .

[gmbcrazy]

@brendonw1 thank you very much. You've already offer so much help, which helps me a lot. I might wait for John's feedback before furtherly bother you, yesterday Peter forward my e-mail to him(author).

Yes blue is raw lfp while red is the filtered. Attached is from another Shank, which is obviously to see the increased ripple power until Chan 21, which might be the one mostly closed to Pyr center. Probably the Channel in .eeg is already arranged from superficial to deep as you said.

[dlevenstein]

Hey @gmbcrazy - is this issue solved?

[gmbcrazy]

@dlevenstein - yes it is. Thanks a lot.