Quantitative sonographic assessment of muscle thickness and fasciculations distribution is a sensitive tool for neuromuscular disorders

PLoS One. 2023 Sep 28;18(9):e0292123. doi: 10.1371/journal.pone.0292123. eCollection 2023.

Abstract

Introduction: Loss of muscle thickness can be demonstrated in a wide spectrum of neuromuscular disorders, while fasciculations are more frequent in amyotrophic lateral sclerosis (ALS). In the current study, we aimed to determine the sensitivity and specificity of quantitative sonographic assessment of muscle thickness and the presence of fasciculations for diagnosing various neuromuscular disorders.

Methods: The thickness and the presence of fasciculations in eight muscles were determined by sonography in patients with myopathy (22), polyneuropathy (36), ALS (91), and spinal muscular atrophy (SMA) (31) and compared to normative values determined in 65 heathy control subjects.

Results: Reduced muscle thickness in at least one relaxed muscle showed 92-100% sensitivity for diagnosing a neuromuscular disease, with a specificity of 85% for differentiating patients from heathy controls (AUC = 0.90). Subtracting distal from proximal muscle thickness may differentiate between myopathy and polyneuropathy. Fasciculations in ≥1 proximal muscle showed good diagnostic accuracy (AUC = 0.87) for diagnosing ALS.

Discussion: Sonographic assessment of muscle thickness is a sensitive tool for diagnosing a wide spectrum of neuromuscular diseases, and may facilitate diagnosis even in patients with normal strength on neurological examination, while the presence of fasciculations in proximal muscles may facilitate ALS diagnosis.

MeSH terms

  • Amyotrophic Lateral Sclerosis* / diagnostic imaging
  • Electromyography
  • Fasciculation / diagnostic imaging
  • Humans
  • Muscle, Skeletal / diagnostic imaging
  • Muscular Diseases*
  • Neuromuscular Diseases* / diagnostic imaging
  • Polyneuropathies* / diagnostic imaging
  • Ultrasonography

Grants and funding

The authors received no specific funding for this work.