skip to main content


Title: Toward a generalizable deep CNN for neural drive estimation across muscles and participants
Abstract

Objective.High-density electromyography (HD-EMG) decomposition algorithms are used to identify individual motor unit (MU) spike trains, which collectively constitute the neural code of movements, to predict motor intent. This approach has advanced from offline to online decomposition, from isometric to dynamic contractions, leading to a wide range of neural-machine interface applications. However, current online methods need offline retraining when applied to the same muscle on a different day or to a different person, which limits their applications in a real-time neural-machine interface. We proposed a deep convolutional neural network (CNN) framework for neural drive estimation, which takes in frames of HD-EMG signals as input, extracts general spatiotemporal properties of MU action potentials, and outputs the number of spikes in each frame. The deep CNN can generalize its application without retraining to HD-EMG data recorded in separate sessions, muscles, or participants.Approach.We recorded HD-EMG signals from the vastus medialis and vastus lateralis muscles from five participants while they performed isometric contractions during two sessions separated by ∼20 months. We identified MU spike trains from HD-EMG signals using a convolutive blind source separation (BSS) method, and then used the cumulative spike train (CST) of these MUs and the HD-EMG signals to train and validate the deep CNN.Main results.On average, the correlation coefficients between CST from the BSS and that from deep CNN were0.983±0.006for leave-one-out across-sessions-and-muscles validation and0.989±0.002for leave-one-out across-participants validation. When trained with more than four datasets, the performance of deep CNN saturated at0.984±0.001for cross validations across muscles, sessions, and participants.Significance.We can conclude that the deep CNN is generalizable across the aforementioned conditions without retraining. We could potentially generate a robust deep CNN to estimate neural drive to muscles for neural-machine interfaces.

 
more » « less
NSF-PAR ID:
10391786
Author(s) / Creator(s):
; ; ; ; ;
Publisher / Repository:
IOP Publishing
Date Published:
Journal Name:
Journal of Neural Engineering
Volume:
20
Issue:
1
ISSN:
1741-2560
Page Range / eLocation ID:
Article No. 016006
Format(s):
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. Abstract

    We present high-precision radial velocities (RVs) from the HARPS-N spectrograph for HD 79210 and HD 79211, two M0V members of a gravitationally bound binary system. We detect a planet candidate with a period of24.4210.017+0.016days around HD 79211 in these HARPS-N RVs, validating the planet candidate originally identified in CARMENES RV data alone. Using HARPS-N, CARMENES, and RVs spanning a total of 25 yr, we further refine the planet candidate parameters toP= 24.422 ± 0.014 days,K= 3.19 ± 0.27 m s−1,Msini= 10.6 ± 1.2M, anda= 0.142 ± 0.005 au. We do not find any additional planet candidate signals in the data of HD 79211, nor do we find any planet candidate signals in HD 79210. This system adds to the number of exoplanets detected in binaries with M-dwarf members and serves as a case study for planet formation in stellar binaries.

     
    more » « less
  2. Abstract

    We present high-resolutionK-band emission spectra of the quintessential hot Jupiter HD 189733 b from the Keck Planet Imager and Characterizer. Using a Bayesian retrieval framework, we fit the dayside pressure–temperature profile, orbital kinematics, mass-mixing ratios of H2O, CO, CH4, NH3, HCN, and H2S, and the13CO/12CO ratio. We measure mass fractions oflogH2O=2.00.4+0.4andlogCO=2.20.5+0.5, and place upper limits on the remaining species. Notably, we find logCH4< −4.5 at 99% confidence, despite its anticipated presence at the equilibrium temperature of HD 189733 b assuming local thermal equilibrium. We make a tentative (∼3σ) detection of13CO, and the retrieved posteriors suggest a12C/13C ratio similar to or substantially less than the local interstellar value. The possible13C enrichment would be consistent with accretion of fractionated material in ices or in the protoplanetary disk midplane. The retrieved abundances correspond to a substantially substellar atmospheric C/O = 0.3 ± 0.1, while the carbon and oxygen abundances are stellar to slightly superstellar, consistent with core-accretion models which predict an inverse correlation between C/O and metallicity. The specific combination of low C/O and high metallicity suggests significant accretion of solid material may have occurred late in the formation process of HD 189733 b.

     
    more » « less
  3. Abstract

    We investigate the stellar mass–black hole mass (*BH) relation with type 1 active galactic nuclei (AGNs) down toBH=107M, corresponding to a ≃ −21 absolute magnitude in rest-frame ultraviolet, atz= 2–2.5. Exploiting the deep and large-area spectroscopic survey of the Hobby–Eberly Telescope Dark Energy Experiment (HETDEX), we identify 66 type 1 AGNs withBHranging from 107–1010Mthat are measured with single-epoch virial method using Civemission lines detected in the HETDEX spectra.*of the host galaxies are estimated from optical to near-infrared photometric data taken with Spitzer, the Wide-field Infrared Survey Explorer, and ground-based 4–8 m class telescopes byCIGALEspectral energy distribution (SED) fitting. We further assess the validity of SED fitting in two cases by host-nuclear decomposition performed through surface brightness profile fitting on spatially resolved host galaxies with the James Webb Space Telescope/NIRCam CEERS data. We obtain the*BHrelation covering the unexplored low-mass ranges ofBH107108M, and conduct forward modeling to fully account for the selection biases and observational uncertainties. The intrinsic*BHrelation atz∼ 2 has a moderate positive offset of 0.52 ± 0.14 dex from the local relation, suggestive of more efficient black hole growth at higher redshift even in the low-mass regime ofBH107108M. Our*BHrelation is inconsistent with theBHsuppression at the low-*regime predicted by recent hydrodynamic simulations at a 98% confidence level, suggesting that feedback in the low-mass systems may be weaker than those produced in hydrodynamic simulations.

     
    more » « less
  4. Abstract

    We present measurements of the rest-frame UV spectral slope,β, for a sample of 36 faint star-forming galaxies atz∼ 9–16 discovered in one of the deepest JWST NIRCam surveys to date, the Next Generation Deep Extragalactic Exploratory Public Survey. We use robust photometric measurements for UV-faint galaxies (down toMUV∼ −16), originally published in Leung et al., and measure values of the UV spectral slope via photometric power-law fitting to both the observed photometry and stellar population models obtained through spectral energy distribution (SED) fitting withBagpipes. We obtain a median and 68% confidence interval forβfrom photometric power-law fitting ofβPL=2.70.5+0.5and from SED fitting,βSED=2.30.1+0.2for the full sample. We show that when only two to three photometric detections are available, SED fitting has a lower scatter and reduced biases than photometric power-law fitting. We quantify this bias and find that after correction the medianβSED,corr=2.50.2+0.2. We measure physical properties for our galaxies withBagpipesand find that our faint (MUV=18.10.9+0.7) sample is low in mass (log[M*/M]=7.70.5+0.5), fairly dust-poor (Av=0.10.1+0.2mag), and modestly young (log[age]=7.80.8+0.2yr) with a median star formation rate oflog(SFR)=0.30.4+0.4Myr1. We find no strong evidence for ultrablue UV spectral slopes (β∼ −3) within our sample, as would be expected for exotically metal-poor (Z/Z< 10−3) stellar populations with very high Lyman continuum escape fractions. Our observations are consistent with model predictions that galaxies of these stellar masses atz∼ 9–16 should have only modestly low metallicities (Z/Z∼ 0.1–0.2).

     
    more » « less
  5. Abstract

    In this work, we develop a deep neural network model for the reaction rate of oxidative coupling of methane from published high-throughput experimental catalysis data. A neural network is formulated so that the rate model satisfies the plug flow reactor design equation. The model is then employed to understand the variation of reactant and product composition within the reactor for the reference catalyst Mn–Na2WO4/SiO2at different temperatures and to identify new catalysts and combinations of known catalysts that would increase yield and selectivity relative to the reference catalyst. The model revealed that methane is converted in the first half of the catalyst bed, while the second part largely consolidates the products (i.e. increases ethylene to ethane ratio). A screening study of3400combinations of pairs of previously studied catalysts of the form M1(M2)12M3Ox/support (where M1, M2 and M3 are metals) revealed that a reactor configuration comprising two sequential catalyst beds leads to synergistic effects resulting in increased yield of C2compared to the reference catalyst at identical conditions and contact time. Finally, an expanded screening study of 7400 combinations (comprising previously studied metals but with several new permutations) revealed multiple catalyst choices with enhanced yields of C2products. This study demonstrates the value of learning a deep neural network model for the instantaneous reaction rate directly from high-throughput data and represents a first step in constraining a data-driven reaction model to satisfy domain information.

     
    more » « less