Questions about HC-2 data usage

Hello,

I have two questions about the hc-2 data. I will take the dataset 'ec013.527' for example.

1.       I have got use LFP signals using the LoadBinary’ script, and the data is a matrix of 31*1325125 elements. On the other hand, the channels is 0:30, what is their relations? And, what is the unit of the data in the matrix?

2.       There are 31 channels in the recording data, but I have got 37 clusters of spikes. How can I assign each cluster to the electrodes.

Thank you very much!

Hi,

1. Channels 0-30 correspond to rows 1-31 in the matlab file. Since the data were recorded originally on 32 electrodes, 1 channel must be removed for this particular file.
1a. The answer to the units is more complicated. For all experiments, the voltage was amplified 20,000x, encoded as a 16-bit signal, usually spanning a range of 10 or 20 V. Thus a value of '1' would correspond to roughly 3.8 to 7.6 picovolts.

2. The spike clusters are unrelated to the number of channels, since the channels correspond to individual electrodes, and the spike clusters are defined by their unique patterns of depolarization measured across electrodes. Thus, one channel  may detect spikes from multiple neurons, which were separated as described in the data description and processing flowchart pdf.

Best,
Gautam

Thanks very much for your reply! But I still have some questions. I hope you can help me, thanks again.

1. I'm still confused about the unit of the LFP waveform in the .eeg files. In the "crcns.org hc2 data description" pdf file, it is mentioned that the neurophysiological signals were amplified 1,000x, but you said  the voltage was amplified 20,000x, why? In addition, taking the data "ec013.527" for example, the voltage range is 20, the maximum value and minimum value is 13441 and -6975 respectively (which are achieved by using the matlab script "LoadBinary"), how can I convert the data point of the LFP waveform to the real neural signal, namely, how can I get the unit? 

2. I've comprehended that the spike clusters are unrelated to the number of channels, but I want to know that how a specified cluster is assigned to the corresponding electrode. In the "LoadCluRes" script, it is commented that "Map is a matrix displaying the correspondance between new cluster numbers (first column) and inital shank # (second column) and cluster # (third column)". In the matrix "Map" of "ec013.527", for example, the 9th line is " 9 1 12", what the "12" means? 

3. In the "crcns.org hc2 data description" pdf file, it is mentioned that the "4-shank or 8-shank silicon probe in layer CA1 of the right dorsal hippocampus". It means that all the electrode are located in the layer CA1, and I cannot obtain the signals in the EC. Is it the fact? If I am right, how can I know which neurons in the deeper parts of the layer and which neurons in the superficial parts of the layer (this is mentioned in the paper "Hippocampal CA1 pyramidal cells form functionally distinct sublayers")? 

Best regards,

Li

Previously Gautam Agarwal wrote:

Hi,

1. Channels 0-30 correspond to rows 1-31 in the matlab file. Since the data were recorded originally on 32 electrodes, 1 channel must be removed for this particular file.
1a. The answer to the units is more complicated. For all experiments, the voltage was amplified 20,000x, encoded as a 16-bit signal, usually spanning a range of 10 or 20 V. Thus a value of '1' would correspond to roughly 3.8 to 7.6 picovolts.

2. The spike clusters are unrelated to the number of channels, since the channels correspond to individual electrodes, and the spike clusters are defined by their unique patterns of depolarization measured across electrodes. Thus, one channel  may detect spikes from multiple neurons, which were separated as described in the data description and processing flowchart pdf.

Best,
Gautam

Hi, I have solved the questions 1 and 3 except for the difference of the amplification. Please reply about the 2nd question and the amplification, thanks!