Journal: Scientific Reports

Article Title: The role of superficial and deep layers in the generation of high frequency oscillations and interictal epileptiform discharges in the human cortex

doi: 10.1038/s41598-022-22497-2

Figure Lengend Snippet: Features of neocortical recording with ripples and IEDs. ( a – f ) CSD and MUA analysis of slow ripples, fast ripples and IEDs. In both cases, changes were considered significant if falling outside the ± 5SD range. MUA values within this range were blurred with green, while significant CSD changes were marked with white contour. ( a ) Cross-patient averaged CSD of slow ripples (80–150 Hz), of the 7 patients with detected HFOs on the neocortex The CSD data represents sink (red) and source (blue) amplitudes. Note the phase reversal in the CSD map in the area of the approximated 2nd cortical layer, and the sudden drop of ripple CSD underneath layer II. Black CSD curve shows averaged waveform from channel 5 (750 μm, Layer II–III border). ( b ) Averaged MUA (500–1000 Hz) corresponding to CSD activity of slow ripples. Unlike the CSD, the MUA showed wider activation during the layer II sink (black frame) including the layers I–II and layer IV–VI. Note a significant high frequency activation in the very bottom layers (arrow) preceding the upper layer MUA activation. The activation phase was followed by an inhibition expanding over wide cortical area, however this did not coincide with layer II source (white frame). ( c ) Averaged CSD pattern of the same n = 7 patients for fast ripple components. Note the alternating sink-source pattern is less self-similar, contain only one full cycle in the average signal. The deeper located sink during the highest amplitude part (black box) is located in the granular layer (layer IV). Corresponding CSD trace is indicated with the black curve in the appropriate layer. ( d ) MUA average for fast ripples. Note the similar deep layer activation compared to slow ripples, ( b ) panel, without the supragranular MUA activation. Also, the post activation decrease is missing. ( e ) Averaged CSD map of the IED discharges of the same n = 7 patients. Black CSD curve is selected from the largest amplitude channel 5 (750 μm, Layer II–III border). It is notable that the alternating sink-source-sink pattern (black frame) extended deeper than the HFO, reaching the layer III. ( f ) Averaged MUA for IED according to panel e. Similarly, to fast ripples, average MUA increase was confined to the infragranular layers (below layer IV). ( g ) Frequency distribution histograms plotted as color coded maps measured for all IEDs across cortical layers. This figure includes all patients with detected IEDs on the neocortex (n = 17). Frequency is binned at 10 Hz precision. Each line of the map corresponds to the recording channels. Approximated cortical layers are indicated on the left-hand side. Note the separation of the distribution pattern at channel 15 (2200 µm cortical depth; approximately the area of layer V). Both domains contain mostly low frequency components (< 150 Hz). ( h ) Frequency distribution measured for all neocortical ripple oscillations (n = 7). Note the upper layer 100–150 Hz, and lower layer (beneath channel 10; 1500 µm cortical depth; approximately layer IV) 200–400 Hz peaks. ( i ) Peak frequency clusters, derived from normalized frequency distribution map at ( g ), using gaussian mixture model (GMM). Peak frequencies are marked with crosses, different colors represent different clusters. Probability distribution of clusters is marked by contours. Mean of probability distribution is indicated by red dots. Horizontal dashed lines highlight the cortical layer coinciding with the mean. The vertical line separates the two clusters at the cutoff frequency value. ( j ) Peak frequency clusters, derived from normalized distribution map at ( h ), using GMM. Note that, compared to IEDs, HFOs present lower cutoff frequency (150 Hz). In addition, however the analysis was optimized to the frequency separation, the mean of the slower HFO cluster falls at higher layers (around layer III), while the faster cluster at somewhat lower, indicating the existence of layer separation of the different frequency components.

Article Snippet: Gaussian mixture model (GMM) with 2 components (fitgmdist and cluster built-in MatLab functions with default parameters) was applied to find frequency-channel clusters in case of both IEDs and HFOs (Fig. i,j).

Techniques: Activity Assay, Activation Assay, Inhibition, Derivative Assay

Journal: Materials

Article Title: Nano-Indentation Properties of Tungsten Carbide-Cobalt Composites as a Function of Tungsten Carbide Crystal Orientation

doi: 10.3390/ma13092137

Figure Lengend Snippet: Parent distributions of both indentation hardness ( a ) and the indentation modulus ( b ). Superimposition of the parent distributions and the individual single/peak Gaussian distributions (blue and red lines), and inherent populations (orange and green bars), associated with rectangular and triangular particles, respectively, in the case of indentation hardness ( c ) and the indentation modulus ( d ). The black dashed line denotes the fitgmdist function trace.

Article Snippet: The MATLAB® fitgmdist built-in function was used and it identified a first single/peak at 22.76 GPa and a second single/peak at 29.61 GPa of 0.84 and 0.16 mixing proportions, respectively.

Techniques: