Impact of functional electrical stimulation on nerve-damaged muscles by quantifying fat infiltration using deep learning

Kassandra Walluks,Jan-Philipp Praetorius,Dirk Arnold,Marc Thilo Figge
DOI: https://doi.org/10.1038/s41598-024-62805-6
IF: 4.6
2024-05-30
Scientific Reports
Abstract:Quantitative imaging in life sciences has evolved into a powerful approach combining advanced microscopy acquisition and automated analysis of image data. The focus of the present study is on the imaging-based evaluation of the posterior cricoarytenoid muscle (PCA) influenced by long-term functional electrical stimulation (FES), which may assist the inspiration of patients with bilateral vocal fold paresis. To this end, muscle cross-sections of the PCA of sheep were examined by quantitative image analysis. Previous investigations of the muscle fibers and the collagen amount have not revealed signs of atrophy and fibrosis due to FES by a laryngeal pacemaker. It was therefore hypothesized that regardless of the stimulation parameters the fat in the muscle cross-sections would not be significantly altered. We here extending our previous investigations using quantitative imaging of intramuscular fat in cross-sections. In order to perform this analysis both reliably and faster than a qualitative evaluation and time-consuming manual annotation, the selection of the automated method was of crucial importance. To this end, our recently established deep neural network IMFSegNet, which provides more accurate results compared to standard machine learning approaches, was applied to more than 300 H&E stained muscle cross-sections from 22 sheep. It was found that there were no significant differences in the amount of intramuscular fat between the PCA with and without long-term FES, nor were any significant differences found between the low and high duty cycle stimulated groups. This study on a human-like animal model not only confirms the hypothesis that FES with the selected parameters has no negative impact on the PCA, but also demonstrates that objective and automated deep learning-based quantitative imaging is a powerful tool for such a challenging analysis.
multidisciplinary sciences
What problem does this paper attempt to address?
The problem that this paper attempts to solve is to evaluate the effect of long - term functional electrical stimulation (FES) on fat infiltration in the impaired neuromuscular system. Specifically, the study focuses on whether functional electrical stimulation will lead to an increase in the fat content in the posterior cricoarytenoid muscle (PCA), which may affect the patient's respiratory function. The study uses quantitative imaging techniques, especially through the deep - learning method (IMFSegNet) to analyze cross - sectional images of H&E - stained muscles to evaluate changes in fat content. ### Main research questions: 1. **The effect of functional electrical stimulation on the fat content in the posterior cricoarytenoid muscle**: The study aims to determine whether long - term functional electrical stimulation will increase the fat content in the posterior cricoarytenoid muscle, especially in the case of using a laryngeal pacemaker (LP) in patients with bilateral vocal cord paralysis (BVFP). 2. **Comparison of quantitative imaging methods**: The study also compares the accuracy and efficiency of different quantitative imaging methods (such as k - means clustering and IMFSegNet) in analyzing the fat content of muscle cross - sections. ### Research background: - **Bilateral vocal cord paralysis (BVFP)**: This disease causes the vocal cords to be unable to open normally, affecting the patient's respiratory function. - **Laryngeal pacemaker (LP)**: It directly stimulates the posterior cricoarytenoid muscle through functional electrical stimulation (FES) to help patients recover their respiratory function. - **The risk of fat infiltration**: Although previous studies have not found that FES causes muscle atrophy or fibrosis, there is a possibility that fat replaces atrophic muscle fibers and collagen, which may affect the function of the muscle. ### Research methods: - **Animal model**: Sheep are used as experimental subjects because their physiological structure is similar to that of humans. - **Image analysis**: Quantitative analysis is carried out through cross - sectional images of H&E - stained muscles, and IMFSegNet is used for automated fat content detection. - **Control group settings**: It includes an unstimulated control group (CT), a sham - operated group with only electrodes implanted but not stimulated (SHAM), and long - term electrical stimulation groups (DC04 and DC07) at different frequencies (12 Hz and 21 Hz). ### Research results: - **No significant difference in fat content**: Whether there is long - term FES or not, there is no significant difference in the fat content in the posterior cricoarytenoid muscle. - **The superiority of IMFSegNet**: Compared with k - means clustering, IMFSegNet performs better in terms of processing time and accuracy and can detect fat content more reliably. - **The potential benefits of electrical stimulation**: Long - term electrical stimulation seems to help reduce fat infiltration caused by nerve damage and electrode implantation. ### Conclusions: - **Safety verification**: Long - term use of the laryngeal pacemaker will not lead to a significant increase in the fat content in the posterior cricoarytenoid muscle, supporting its safe clinical application. - **Technology promotion**: The application of deep - learning methods (such as IMFSegNet) in quantitative imaging not only improves the accuracy and efficiency of analysis but also provides new tools for clinical diagnosis. Through these studies, the author provides strong evidence for the clinical application of laryngeal pacemakers in patients with bilateral vocal cord paralysis and shows the great potential of deep learning in biomedical image analysis.