Journal: Current Biology

Article Title: Local and system mechanisms for action execution and observation in parietal and premotor cortices

doi: 10.1016/j.cub.2021.04.034

Figure Lengend Snippet: Identification and functional properties of cell classes based on extracellular spike waveforms (A) Projection of each spike waveform in the 2D space formed by trough-to-peak duration and repolarization time. Color codes identify the clusters (cell classes) resulting from the Gaussian mixture model applied with the number of components (n = 3) indicated by the Bayesian information criterion (BIC) shown in the inset . The black dots in each cluster indicate the example neurons shown in (D). Colored ellipses indicate, for each cluster, the 2D confidence interval. Trough-to-peak values range from 0.13 ms to 0.58 ms, and repolarization time values range from 0.0025 ms to 0.43 ms. Average variability in trough-peak estimation is 3.1 μs (95 th percentile = 7.9 μs); average variability in repolarization time estimation is 14.8 μs (95 th percentile = 65.4 μs). See Figure S2 for clustering reliability within and across areas. Figure S3 A shows alternative clustering results obtained using spiking and waveform features. (B) Separation among cell classes. For each of 10 4 data points randomly generated from the fitted Gaussian mixture distribution, we compared the true class from which the point was drawn with the class to which it was assigned. The confusion matrix shows the classification results; accuracy is 0.95 and results from the mean of the three diagonal probabilities. (C) Number of neurons in each cell class (in color code) in the entire dataset and individual average spike waveforms belonging to each class. (D) Example neurons recorded in AIP, F5, and F6 (from Neurons 1 to 3; see black circles in A), belonging to each of the three classes (spike waveform is shown in the inset of each histogram; color code as in B). Activity is aligned (vertical dashed lines) on object presentation (Obj pres) and then (after the gap) on the Go signal, in both tasks. Each color refers to trials with one type of target object: a ring (red), a small cone (blue), and a big cone (black). Triangular markers indicate the movement onset (green) and object pulling onset (yellow).

Article Snippet: Then, to identify clusters of waveforms based on these two parameters, we followed a recently described procedure in which the two-dimensional data points are fitted with a Gaussian mixture distribution (MATLAB function: fitgmdist).

Techniques: Functional Assay, Generated, Activity Assay

Journal: Frontiers in Public Health

Article Title: French Version of the Antiphasic Digits-in-Noise Test for Smartphone Hearing Screening

doi: 10.3389/fpubh.2021.725080

Figure Lengend Snippet: Höra screening campaign results. (A) Shown is the age repartition of the Höra population (green, left axis) compared to the reference population (gray, right axis). Curves correspond to the sum of two gaussian model (reference population, gray, f ( x ) = 3.76 × exp(–((x − 26)/6.94) 2 ) + 3.66 × exp(–((x − 67.8)/18.5) 2 ), R 2 = 0.72, Höra population, green, f ( x ) = 668.46 × exp(–((x − 20)/6.24) 2 ) + 332 × exp(–((x − 78.6)/20.6) 2 , R 2 = 0.95). (B) Mean ± SEM of digits-in-noise speech reception thresholds for reference (gray) and Höra (green) population relative to age and the corresponding broken-stick regression. The inserted graph presents the cut-off values of broken-stick regression for reference (gray bar) and Höra (blue bar). Note that the cut-off age for PTA (44 years-old) and DIN SRT (46 years-old) are not statistically different. (C) Shown is the DIN SRT distribution across all the subjects participating to the Höra screening campaign (green). The inserted graph presents the distribution of Höra tests as normal hearing (74%), mild hearing loss (9%), or moderate or worst hearing loss (17%).

Article Snippet: Distribution of results were calculated and fitted with unimodal (one) or bimodal (two) Gaussian models (Matlab functions: hist, fitgmdist).

Techniques: