Intra-burst firing characteristics as network state parameters
Stegenga, Jan and Feber le, Joost and Rutten, Wim and Marani, Enrico (2006) Intra-burst firing characteristics as network state parameters. In: 5th international meeting on substrate-integrated micro electrode arrays. BIOPRO Baden-Wuerttemberg, Stuttgart, pp. 64-66. ISBN 9783938345023
In our group we are aiming to demonstrate learning and memory capabilities of cultured networks of cortical neurons. A first step is to identify parameters that accurately describe changes in the network due to learning. Usually, such parameters are calculated from the responses to test-stimuli before and after a learning experiment. We propose that parameters should be calculated from the spontaneous activity before and after a learning experiment, as the applying of test-stimuli itself may alter the network. Since bursting is dominant in our cultures, we have investigated its spatio-temporal structure.
Networks of cortical neurons were cultured on a MEA. Over a period from 9 to 35 DIV, the spontaneous activity has been measured on a regular basis. Measurements on a single day are always continuous; otherwise cultures are kept in a stove under controlled conditions (37 ˚C, 5% CO2, 100% humidity). Network bursts were detected by analysing the Array-Wide Spiking Rate (AWSR, the sum of activity over all electrodes). Next, we estimated the instantaneous AWSR during a burst by convolving spike-occurrences with a Gaussian function. We investigated the changes in burst profiles over time by aligning them to their peak AWSR. In 4 hour recording sessions, we grouped the burst profiles over 1 hour, resulting in 4 average burst profiles per day. In addition, a sufficient amount of aligned bursts yielded enough data to calculate the contribution of each recording site.
The burst profiles, calculated over a period of 1 hour, generally show little variation (figure 1). In subsequent hours, the profiles gradually change shape. Over a period of days however, the shape can change dramatically (figure 2). The relatively slow changes over the period of hours indicate an underlying probabilistic structure in the AWSR during bursts. The apparent structure in the burst profiles result from the relationships between individual recording sites, and thus also on the connectivity in the neural network. This is revealed in more detail by showing the contributions of individual sites (figure 3). The spike envelopes have a shape that is too detailed to be described accurately by a small set of parameters.
The burst profiles prove to be stable over a period of one hour, and gradually change their shape over several hours, as has also been suggested in . The day-to-day changes in burst profiles may be the result of these gradual changes, thereby suggesting an intrinsically changing network. However, they can also be the result of putting the cultures back in the stove. The spike envelopes per recording site offer more detailed descriptions of the network state than the burst profiles. This may however be the amount of detail required to reveal the changes made during learning experiments. A subsequent refinement can be made by identifying distinct subgroups of bursts, as has been suggested in .
|Item Type:||Book Section|
Electrical Engineering, Mathematics and Computer Science (EEMCS)
|Link to this item:||http://purl.utwente.nl/publications/60467|
|Export this item as:||BibTeX|
Daily downloads in the past month
Monthly downloads in the past 12 months
Repository Staff Only: item control page
Metis ID: 237628