Variability Of Acoustic Properties Research Articles

Data-driven machine learning models for decoding speech categorization from evoked brain responses

Objective. Categorical perception (CP) of audio is critical to understand how the human brain perceives speech sounds despite widespread variability in acoustic properties. Here, we investigated the spatiotemporal characteristics of auditory neural activity that reflects CP for speech (i.e. differentiates phonetic prototypes from ambiguous speech sounds). Approach. We recorded 64-channel electroencephalograms as listeners rapidly classified vowel sounds along an acoustic-phonetic continuum. We used support vector machine classifiers and stability selection to determine when and where in the brain CP was best decoded across space and time via source-level analysis of the event-related potentials. Main results. We found that early (120 ms) whole-brain data decoded speech categories (i.e. prototypical vs. ambiguous tokens) with 95.16% accuracy (area under the curve 95.14%; F1-score 95.00%). Separate analyses on left hemisphere (LH) and right hemisphere (RH) responses showed that LH decoding was more accurate and earlier than RH (89.03% vs. 86.45% accuracy; 140 ms vs. 200 ms). Stability (feature) selection identified 13 regions of interest (ROIs) out of 68 brain regions [including auditory cortex, supramarginal gyrus, and inferior frontal gyrus (IFG)] that showed categorical representation during stimulus encoding (0–260 ms). In contrast, 15 ROIs (including fronto-parietal regions, IFG, motor cortex) were necessary to describe later decision stages (later 300–800 ms) of categorization but these areas were highly associated with the strength of listeners’ categorical hearing (i.e. slope of behavioral identification functions). Significance. Our data-driven multivariate models demonstrate that abstract categories emerge surprisingly early (∼120 ms) in the time course of speech processing and are dominated by engagement of a relatively compact fronto-temporal-parietal brain network.

Acoustic Properties of Ancient Shallow-Marine Carbonates: Effects of Depositional Environments and Diagenetic Processes (Middle Jurassic, Paris Basin, France)

Abstract Examination of petrophysical properties (acoustic velocity, porosity, permeability, and density) and petrographical characteristics (texture, facies composition, and diagenesis) of more than 250 core plugs from the Middle Jurassic carbonates of the eastern Paris Basin provides insights into the parameters controlling acoustic velocities in relatively low-porosity carbonate rocks (Φ < 20%). The pore-type observations reveal distinct acoustic velocities in samples with intergranular macropores and samples with micropores in subhedral micrite, such that velocities in microporous mudstone–wackestone (lagoonal) deposits are greater than in macroporous grainstone (shoal) samples, at a given porosity range (15–20%). The standard Wyllie and Raymer transforms fit very well with the linear regression between acoustic velocity and porosity from mudstone or lagoonal facies. Marls and fine-grained deposits interpreted as lagoonal facies include statistically significant correlation (r = 0.9) between velocity and porosity. However, the data suggest that the wide scatter in velocity–porosity relationship from grainstones are not the result of different sorting, grain size, pore type, dolomite content, or clay content. Instead, early cementation greatly influences acoustic properties during diagenesis, and are interpreted to account for the high variability of velocities over a given porosity range. Specifically, at a given porosity, acoustic velocities in compacted grainstone that did not undergo early cementation are higher than in early-cemented grainstone. Petrographic observations suggest that early cementation limits mechanical compaction, creating a heterogeneous medium from the earliest stages of diagenesis (non-touching grains, preservation of intergranular macropores that are partially to totally filled by later blocky calcite cement). The abundant interfaces between micritized ooids, early cement fringes, and blocky calcites in grainstones may induce significant wave attenuation. As a result, the standard time-average equations fail to predict the effect of diagenetic features such as early cementation on sonic velocity. Conversely, an absence of early cementation favors mechanical compaction, grain-to-grain contact, and suturing. The result is a homogeneous micritized grain-supported network that may facilitate wave propagation. Through demonstration of the key role of early cementation in the explanation of variability in acoustic properties, the results of this study illustrate the complicated factors influencing velocity transforms in carbonates (Wyllie and Raymer), i.e., classical tools for predicting reservoir properties. These insights on the interpretation of Vp and the refinement of velocity–porosity transforms in grainstone units may be broadly applicable to enhancing seismic-based exploration in carbonate successions.