Assessing spinal muscular atrophy with quantitative ultrasound

Machine learning algorithms to classify spinal muscular atrophy subtypes

OBJECTIVES

The development of better biomarkers for disease assessment remains an ongoing effort across the spectrum of neurologic illnesses. One approach for refining biomarkers is based on the concept of machine learning, in which individual, unrelated biomarkers are simultaneously evaluated. In this cross-sectional study, we assess the possibility of using machine learning, incorporating both quantitative muscle ultrasound (QMU) and electrical impedance myography (EIM) data, for classification of muscles affected by spinal muscular atrophy (SMA).

METHODS

Twenty-one normal subjects, 15 subjects with SMA type 2, and 10 subjects with SMA type 3 underwent EIM and QMU measurements of unilateral biceps, wrist extensors, quadriceps, and tibialis anterior. EIM and QMU parameters were then applied in combination using a support vector machine (SVM), a type of machine learning, in an attempt to accurately categorize 165 individual muscles.

RESULTS

For all 3 classification problems, normal vs SMA, normal vs SMA 3, and SMA 2 vs SMA 3, use of SVM provided the greatest accuracy in discrimination, surpassing both EIM and QMU individually. For example, the accuracy, as measured by the receiver operating characteristic area under the curve (ROC-AUC) for the SVM discriminating SMA 2 muscles from SMA 3 muscles was 0.928; in comparison, the ROC-AUCs for EIM and QMU parameters alone were only 0.877 (p < 0.05) and 0.627 (p < 0.05), respectively.

CONCLUSIONS

Combining EIM and QMU data categorizes individual SMA-affected muscles with very high accuracy. Further investigation of this approach for classifying and for following the progression of neuromuscular illness is warranted.

New therapeutic approaches to spinal muscular atrophy

Bench to bedside progress has been widely anticipated for a growing number of neurodegenerative disorders. Of these, spinal muscular atrophy (SMA) is perhaps the best poised to capitalize on advances in targeted therapeutics development over the next few years. Several laboratories have achieved compelling success in SMA animal models using sophisticated methods for targeted delivery, repair, or increased expression of the survival motor neuron protein, SMN. The clinical community is actively collaborating to identify, develop, and validate outcome measures and biomarkers in parallel with laboratory efforts. Innovative trial design and synergistic approaches to maximize proactive care in conjunction with treatment with one or more of the promising pharmacologic and biologic therapies currently in the pipeline will maximize our chances to achieve meaningful outcomes for patients. This review highlights recent promising scientific and clinical advances bringing us ever closer to effective treatment(s) for our patients with SMA.

Ultrasound is a reliable measure of muscle thickness in acute stroke patients, for some, but not all anatomical sites: a study of the intra-rater reliability of muscle thickness measures in acute stroke patients

This prospective, blinded study investigates the test retest reliability of measures of muscle thickness made by one sonographer across two cohort groups (n = 29) of people hospitalised with acute stroke. Reliability was assessed in cohort one (n = 14) for measurements made bilaterally at the anterior and posterior upper arms, the anterior and posterior thighs (total of eight measurements) and in cohort two (n = 15), for measurements made bilaterally at the lateral forearms, the anterior abdominal wall and the anterior and lower legs (total of eight measurements). Reliability estimates varied between measurement sites; intraclass correlation coefficients (ICCs) ranged from -0.26 (lateral forearm, paretic side) to 0.95 (anterior thigh, nonparetic side), percent mean differences ranged from 0.42% (posterior upper arm, nonparetic side) to 14.68% (anterior lower limb, nonparetic side) and method error ranged from 1.08 (abdomen, nonparetic side) to 9.69 mm (posterior lower limb, nonparetic side). Only four measurement sites (anterior upper arm, posterior upper arm, abdomen and anterior thigh) were within the acceptable ranges (ICC 0.60 to 1.00, mean percent difference range 0%-5% and method error range 0-5 mm) and considered reliable to use for measures of muscle thickness in people hospitalised with acute stroke.

Spinal muscular atrophy: a clinical and research update

Spinal muscular atrophy, a hereditary degenerative disorder of lower motor neurons associated with progressive muscle weakness and atrophy, is the most common genetic cause of infant mortality. It is caused by decreased levels of the "survival of motor neuron" (SMN) protein. Its inheritance pattern is autosomal recessive, resulting from mutations involving the SMN1 gene on chromosome 5q13. However, unlike many other autosomal recessive diseases, the SMN gene involves a unique structure (an inverted duplication) that presents potential therapeutic targets. Although no effective treatment for spinal muscular atrophy exists, the field of translational research in spinal muscular atrophy is active, and clinical trials are ongoing. Advances in the multidisciplinary supportive care of children with spinal muscular atrophy also offer hope for improved life expectancy and quality of life.

Neuromuscular ultrasonography: quantifying muscle and nerve measurements

Neuromuscular ultrasonography can be used both descriptively and quantitatively in the evaluation of patients with neuromuscular disorders. This article reviews the quantitative use of this technology, particularly measurements of the size and echogenicity of nerve and muscle, as a tool for assessing the severity, progression, and response of these tissues to therapeutic interventions. Neuromuscular ultrasonography has several features, including portability and noninvasiveness, that make it an attractive research tool for advancing the diagnosis and treatment of neuromuscular disorders.

Peripheral nerve and muscle ultrasound in amyotrophic lateral sclerosis

INTRODUCTION

High-resolution ultrasound has been used to evaluate several neuromuscular conditions, but it has only been used on a limited basis in amyotrophic lateral sclerosis (ALS) patients. It has not been used to assess their peripheral nerves. This study was designed to use neuromuscular ultrasound to investigate nerve cross-sectional area and muscle thickness in ALS.

METHODS

Twenty individuals with ALS and 20 matched controls underwent neuromuscular ultrasound to measure the cross-sectional area of their median and sural nerves and the thickness of their biceps/brachialis muscle complex.

RESULTS

The cross-sectional area of the median nerve in the mid-arm was smaller in the ALS group than in controls (10.5 mm(2) vs. 12.7 mm(2) , P = 0.0023), but no difference was seen in the sural nerve (4.5 mm(2) vs. 5.0 mm(2) , P = 0.1927). The ALS group also had thinner biceps/brachialis than controls (2.1 cm vs. 2.9 cm, P = 0.0007).

CONCLUSIONS

Neuromuscular ultrasound demonstrates nerve and muscle atrophy in ALS and should be further explored as a disease biomarker.