A fast blind source separation algorithm for decomposing ultrafast ultrasound images into spatiotemporal muscle unit kinematics

Abstract

Objective: Ultrasound can detect individual motor unit (MU) activity during voluntary isometric contractions based on their subtle axial displacements. The detection pipeline, currently performed offline, is based on displacement velocity images and identifying the subtle axial displacements. This identification can preferably be made through a blind source separation (BSS) algorithm with the feasibility of translating the pipeline from offline to online. However, the question remains how to reduce the computational time for the BSS algorithm, which includes demixing tissue velocities from many different sources, e.g., the active MU displacements, arterial pulsations, bones, connective tissue, and noise.

Approach: This study proposes a fast velocity-based BSS (velBSS) algorithm suitable for online purposes that decomposes velocity images from low-force voluntary isometric contractions into spatiotemporal components associated with single MU activities. The proposed algorithm will be compared against spatiotemporal independent component analysis (stICA), i.e., the method used in previous papers, for various subjects, ultrasound- and EMG systems, where the latter acts as MU reference recordings.

Main results: We found that the computational time for velBSS was at least 20 times less than for stICA, while the twitch responses and spatial maps extracted from stICA and velBSS for the same MU reference were highly correlated (0.96 ± 0.05 and 0.81 ± 0.13).

Significance: The present algorithm (velBSS) is computationally much faster than the currently available method (stICA) while maintaining the same performance. It provides a promising translation towards an online pipeline and will be important in the continued development of this research field of functional neuromuscular imaging.

Spatially repeatable components from ultrafast ultrasound are associated with motor unit activity in human isometric contractions *

“…Therefore, the blind source separation approach provides advantages compared to the spiketriggered averaging approach, such as lower memory and storage requirements and potential for, e.g. realtime imaging [41] and dynamic contractions applications. For these applications, future studies must consider the lower bound in terms of the recording duration to identify MUs and improve the classification of components into MUs or non-MUs using robust features or training a classifier.…”

Spatial decomposition of ultrafast ultrasound images to identify motor unit activity – A validation study using intramuscular and surface EMG

“…Also, the algorithms decoding performance of velocity images are unknown. Therefore, we used a fast spatial BSS method based on a linear model (Rohlén et al, 2023c) similar to the ones used in previous studies (Carbonaro et al, 2022;Rohlén et al, 2020a).…”

“…First, the velocity image sequence 𝒀 were z-scored along the time dimension, and the mean was subtracted in both dimensions. Second, random SVD (with exponent parameter 𝑞 = 1 and oversampling parameter 𝑝 = 5 (Halko et al, 2011;Rohlén et al, 2023c)) was applied to decompose 𝒀 into 𝑘 = 100 (Rohlén et al, 2020b) left and right singular vectors. Third, 𝒀 were whitened (Hyvärinen et al, 2001) based on the left and right singular vectors.…”

“…MU decoding from UUS has been attempted with blind source separation (BSS) (Rohlén et al, 2020b(Rohlén et al, , 2020a. This approach is computationally fast and suitable for real-time processing (Rohlén et al, 2023c). However, the BSS method has yet to be validated for a large population of MUs.…”

“…This study aimed to validate the BSS method (Rohlén et al, 2023c) to identify individual MU activity in UUS recordings using a large population of MUs. The validation was achieved by assessing its ability to identify firings, twitch areas and temporal profiles in the tibialis anterior (TA) of eleven healthy participants with concurrent recordings of either high-density surface EMG (HDsEMG) or fine-wire intramuscular EMG (iEMG).…”