Estimation of absolute states of human skeletal muscle via standard B-mode ultrasound imaging and deep convolutional neural networks

The objective is to test automated in vivo estimation of active and passive skeletal muscle states using ultrasonic imaging. Current technology (electromyography, dynamometry, shear wave imaging) provides no general, non-invasive method for online estimation of skeletal muscle states. Ultrasound (US) allows non-invasive imaging of muscle, yet current computational approaches have never achieved simultaneous extraction or generalization of independently varying active and passive states. We use deep learning to investigate the generalizable content of two-dimensional (2D) US muscle images. US data synchronized with electromyography of the calf muscles, with measures of joint moment/angle, were recorded from 32 healthy participants (seven female; ages: 27.5, 19–65). We extracted a region of interest of medial gastrocnemius and soleus using our prior developed accurate segmentation algorithm. From the segmented images, a deep convolutional neural network was trained to predict three absolute, drift-free components of the neurobiomechanical state (activity, joint angle, joint moment) during experimentally designed, simultaneous independent variation of passive (joint angle) and active (electromyography) inputs. For all 32 held-out participants (16-fold cross-validation) the ankle joint angle, electromyography and joint moment were estimated to accuracy 55 ± 8%, 57 ± 11% and 46 ± 9%, respectively. With 2D US imaging, deep neural networks can encode, in generalizable form, the activity–length–tension state relationship of these muscles. Observation-only, low-power 2D US imaging can provide a new category of technology for non-invasive estimation of neural output, length and tension in skeletal muscle. This proof of principle has value for personalized muscle assessment in pain, injury, neurological conditions, neuropathies, myopathies and ageing.

Download Full-text

The Application of Deep Convolutional Neural Networks to Ultrasound for Modelling of Dynamic States within Human Skeletal Muscle

10.1101/157479 ◽

2017 ◽

Cited By ~ 2

Author(s):

Ryan J. Cunningham ◽

Peter J. Harding ◽

Ian D. Loram

Keyword(s):

Skeletal Muscle ◽

Joint Angle ◽

Region Of Interest ◽

Image Understanding ◽

Joint Torque ◽

Medial Gastrocnemius ◽

Ultrasound Images ◽

Deep Convolutional Neural Networks ◽

Segmented Images

AbstractThis paper concerns the fully automatic direct in vivo measurement of active and passive dynamic skeletal muscle states using ultrasound imaging. Despite the long standing medical need (myopathies, neuropathies, pain, injury, ageing), currently technology (electromyography, dynamometry, shear wave imaging) provides no general, non-invasive method for online estimation of skeletal intramuscular states. Ultrasound provides a technology in which static and dynamic muscle states can be observed non-invasively, yet current computational image understanding approaches are inadequate. We propose a new approach in which deep learning methods are used for understanding the content of ultrasound images of muscle in terms of its measured state. Ultrasound data synchronized with electromyography of the calf muscles, with measures of joint torque/angle were recorded from 19 healthy participants (6 female, ages: 30 ± 7.7). A segmentation algorithm previously developed by our group was applied to extract a region of interest of the medial gastrocnemius. Then a deep convolutional neural network was trained to predict the measured states (joint angle/torque, electromyography) directly from the segmented images. Results revealed for the first time that active and passive muscle states can be measured directly from standard b-mode ultrasound images, accurately predicting for a held out test participant changes in the joint angle, electromyography, and torque with as little error as 0.022°, 0.0001V, 0.256Nm (root mean square error) respectively.

Download Full-text

Introduction to in vivo31P magnetic resonance spectroscopy of (human) skeletal muscle

Proceedings of The Nutrition Society ◽

10.1017/s0029665199001160 ◽

1999 ◽

Vol 58 (4) ◽

pp. 861-870 ◽

Cited By ~ 32

Author(s):

A. Heerschap ◽

C. Houtman ◽

H. J. A. in 't Zandt ◽

A. J. van den Bergh ◽

B. Wieringa

Keyword(s):

Skeletal Muscle ◽

Magnetic Resonance ◽

Magnetic Resonance Spectroscopy ◽

Current Status ◽

Resonance Spectroscopy ◽

Biochemical Processes ◽

Non Invasive ◽

Introductory Overview ◽

Muscle Biochemistry

31P magnetic resonance spectroscopy (MRS) offers a unique non-invasive window on energy metabolism in skeletal muscle, with possibilities for longitudinal studies and of obtaining important bioenergetic data continuously and with sufficient time resolution during muscle exercise. The present paper provides an introductory overview of the current status of in vivo31P MRS of skeletal muscle, focusing on human applications, but with some illustrative examples from studies on transgenic mice. Topics which are described in the present paper are the information content of the 31P magnetic resonance spectrum of skeletal muscle, some practical issues in the performance of this MRS methodology, related muscle biochemistry and the validity of interpreting results in terms of biochemical processes, the possibility of investigating reaction kinetics in vivo and some indications for fibre-type heterogeneity as seen in spectra obtained during exercise.

Download Full-text

Optogenetic activation of muscle contraction in vivo

10.1101/2020.07.07.192377 ◽

2020 ◽

Author(s):

Elahe Ganji ◽

C. Savio Chan ◽

Christopher W. Ward ◽

Megan L. Killian

Keyword(s):

Skeletal Muscle ◽

Electrical Stimulation ◽

Muscle Contraction ◽

Fluorescent Imaging ◽

Triceps Surae ◽

Light Activation ◽

Non Invasive ◽

Optogenetic Stimulation ◽

Stimulation Of

AbstractOptogenetics is an emerging alternative to traditional electrical stimulation to initiate action potentials in activatable cells both ex vivo and in vivo. Optogenetics has been commonly used in mammalian neurons and more recently, it has been adapted for activation of cardiomyocytes and skeletal muscle. Therefore, the aim of this study was to evaluate the stimulation feasibility and sustain isometric muscle contraction and limit decay for an extended period of time (1s), using non-invasive transdermal light activation of skeletal muscle (triceps surae) in vivo. We used inducible Cre recombination to target expression of Channelrhodopsin-2 (ChR2(H134R)-EYFP) in skeletal muscle (Acta1-Cre) in mice. Fluorescent imaging confirmed that ChR2 expression is localized in skeletal muscle and does not have specific expression in sciatic nerve branch, therefore, allowing for non-nerve mediated optical stimulation of skeletal muscle. We induced muscle contraction using transdermal exposure to blue light and selected 10Hz stimulation after controlled optimization experiments to sustain prolonged muscle contraction. Increasing the stimulation frequency from 10Hz to 40Hz increased the muscle contraction decay during prolonged 1s stimulation, highlighting frequency dependency and importance of membrane repolarization for effective light activation. Finally, we showed that optimized pulsed optogenetic stimulation of 10 Hz resulted in comparable ankle torque and contractile functionality to that of electrical stimulation. Our results demonstrate the feasibility and repeatability of non-invasive optogenetic stimulation of muscle in vivo and highlight optogenetic stimulation as a powerful tool for non-invasive in vivo direct activation of skeletal muscle.

Download Full-text

A Vision-Based Method Utilizing Deep Convolutional Neural Networks for Fruit Variety Classification in Uncertainty Conditions of Retail Sales

Applied Sciences ◽

10.3390/app9193971 ◽

2019 ◽

Vol 9 (19) ◽

pp. 3971 ◽

Cited By ~ 4

Author(s):

Katarzyna ◽

Paweł

Keyword(s):

Neural Networks ◽

Convolutional Neural Networks ◽

Region Of Interest ◽

Certainty Factor ◽

Retail Sales ◽

Deep Convolutional Neural Networks ◽

Testing Dataset ◽

Variety Classification ◽

Sales Systems ◽

Double Track

This study proposes a double-track method for the classification of fruit varieties for application in retail sales. The method uses two nine-layer Convolutional Neural Networks (CNNs) with the same architecture, but different weight matrices. The first network classifies fruits according to images of fruits with a background, and the second network classifies based on images with the ROI (Region Of Interest, a single fruit). The results are aggregated with the proposed values of weights (importance). Consequently, the method returns the predicted class membership with the Certainty Factor (CF). The use of the certainty factor associated with prediction results from the original images and cropped ROIs is the main contribution of this paper. It has been shown that CFs indicate the correctness of the classification result and represent a more reliable measure compared to the probabilities on the CNN outputs. The method is tested with a dataset containing images of six apple varieties. The overall image classification accuracy for this testing dataset is excellent (99.78%). In conclusion, the proposed method is highly successful at recognizing unambiguous, ambiguous, and uncertain classifications, and it can be used in a vision-based sales systems in uncertain conditions and unplanned situations.

Download Full-text

In-vivo 31 P-MRS of skeletal muscle and liver: A way for non-invasive assessment of their metabolism

Analytical Biochemistry ◽

10.1016/j.ab.2017.01.018 ◽

2017 ◽

Vol 529 ◽

pp. 193-215 ◽

Cited By ~ 35

Author(s):

Ladislav Valkovič ◽

Marek Chmelík ◽

Martin Krššák

Keyword(s):

Skeletal Muscle ◽

Non Invasive

Download Full-text

Improving representations of genomic sequence motifs in convolutional networks with exponential activations

10.1101/2020.06.14.150706 ◽

2020 ◽

Cited By ~ 3

Author(s):

Peter K. Koo ◽

Matt Ploenzke

Keyword(s):

Neural Networks ◽

Dna Sequences ◽

Test Performance ◽

Genomic Sequence ◽

Comprehensive Analysis ◽

Sequence Motifs ◽

Deep Convolutional Neural Networks ◽

Convolutional Networks ◽

Learned Features

ABSTRACTDeep convolutional neural networks (CNNs) trained on regulatory genomic sequences tend to build representations in a distributed manner, making it a challenge to extract learned features that are biologically meaningful, such as sequence motifs. Here we perform a comprehensive analysis on synthetic sequences to investigate the role that CNN activations have on model interpretability. We show that employing an exponential activation to first layer filters consistently leads to interpretable and robust representations of motifs compared to other commonly used activations. Strikingly, we demonstrate that CNNs with better test performance do not necessarily imply more interpretable representations with attribution methods. We find that CNNs with exponential activations significantly improve the efficacy of recovering biologically meaningful representations with attribution methods. We demonstrate these results generalise to real DNA sequences across several in vivo datasets. Together, this work demonstrates how a small modification to existing CNNs, i.e. setting exponential activations in the first layer, can significantly improve the robustness and interpretabilty of learned representations directly in convolutional filters and indirectly with attribution methods.

Download Full-text

UFaceNet: Research on Multi-Task Face Recognition Algorithm Based on CNN

Algorithms ◽

10.3390/a14090268 ◽

2021 ◽

Vol 14 (9) ◽

pp. 268

Author(s):

Huoyou Li ◽

Jianshiun Hu ◽

Jingwen Yu ◽

Ning Yu ◽

Qingqiang Wu

Keyword(s):

Neural Networks ◽

Face Recognition ◽

Recognition Algorithm ◽

Training Data ◽

Deep Convolutional Neural Networks ◽

Model Calculations ◽

Keypoint Detection ◽

In Vivo Detection ◽

The Face

With the application of deep convolutional neural networks, the performance of computer vision tasks has been improved to a new level. The construction of a deeper and more complex network allows the face recognition algorithm to obtain a higher accuracy, However, the disadvantages of large computation and storage costs of neural networks limit the further popularization of the algorithm. To solve this problem, we have studied the unified and efficient neural network face recognition algorithm under the condition of a single camera; we propose that the complete face recognition process consists of four tasks: face detection, in vivo detection, keypoint detection, and face verification; combining the key algorithms of these four tasks, we propose a unified network model based on a deep separable convolutional structure—UFaceNet. The model uses multisource data to carry out multitask joint training and uses the keypoint detection results to aid the learning of other tasks. It further introduces the attention mechanism through feature level clipping and alignment to ensure the accuracy of the model, using the shared convolutional layer network among tasks to reduce model calculations amount and realize network acceleration. The learning goal of multi-tasking implicitly increases the amount of training data and different data distribution, making it easier to learn the characteristics with generalization. The experimental results show that the UFaceNet model is better than other models in terms of calculation amount and number of parameters with higher efficiency, and some potential areas to be used.

Download Full-text

NON-INVASIVE ANALYSIS OF THE MECHANICAL PROPERTIES OF SKELETAL MUSCLE TISSUE IN-VIVO

10.31224/osf.io/mejbx ◽

2019 ◽

Author(s):

Kevin Mattheus Moerman ◽

Andre Sprengers ◽

Aart Nederveen ◽

Ciaran Simms

Keyword(s):

Skeletal Muscle ◽

Muscle Tissue ◽

Constitutive Models ◽

Skeletal Muscle Tissue ◽

Constitutive Law ◽

Element Analysis ◽

Indentation Force ◽

Non Invasive ◽

Inverse Finite Element Analysis

Validation of constitutive models of living human skeletal muscle tissue requires non-invasive analysis methods. This abstract presents an advanced framework for the validation of such models. MRI based indentation experiments are proposed during which dynamic indentation force, 3D dynamic tissue deformation, tissue geometry and fibre architecture are recorded. Following inverse Finite Element Analysis (FEA) of the experimental boundary conditions the parameters of an appropriate constitutive law can be optimised. The framework presented allows for the first time for the non-invasive evaluation of detailed non-linear, anisotropic and viscoelastic material laws.

Download Full-text