Validation of Neuromyotype: a smart keyboard for the evaluation of spinal muscular atrophy patients

Abstract

Introduction

Spinal muscular atrophy 5q (SMA) is a genetic neurodegenerative disease that affects alpha motor neurons producing progressive weakness. New outcome measures are currently required to accurately characterise the disease progression and the efficacy of new available treatments. The objective of this work is to preliminarily validate a new intelligent keyboard (Neuromyotype) measuring typing strength and speed in patients with SMA.

Material and methods

Twenty two SMA patients older than 15 years, and 26 healthy controls were included. Three measurements were obtained with the keyboard (maximum strength, execution time of a random typing task, execution time of a sequential typing task) together with the time to complete the Nine-Hole Peg Test (9HPT). Patients were also administered motor (Hammersmith Functional Motor Scale Expanded, HFMSE; Revised Upper Limb module, RULM), and functional scales (Egen Klassification, EK2; and the revised version of Amyotrophic Lateral Sclerosis Functional Rating Scale, ALSFRS-R). The viability and construct validity of the Neuromyotype were analysed, measuring the discriminative power between patients and controls (using ROC curves and the Bangdiwala's B statistic), between the different functional types of SMA (walker, sitter and non-sitter) and their correlation with the rest of motor scales.

Results

Neuromyotype measurements could be performed in all patients, unlike the rest of the scales. Its administration was quick and easy. The 3 variables on the keyboard discriminated very well between patients and controls, with strength (ROC = 0.963) being the one that best differentiates from the 3, equaling 9HPT (ROC = 0.966). They also showed a good ability to differentiate by functional type (especially non-sitters from sitters and walkers), with sequential time (B = 0.83) being the tool that best discriminates between the three groups above the rest of motor scales. All motor and functional scales showed strong or very strong correlations with each other (rs = 0.71–0.99), with strength correlating better with motor scales and timed variables with functional scales.

Conclusion

This study shows the feasibility and validity of Neuromyotype for the evaluation of adolescent and adult patients with SMA. Data obtained with this tool could be of great clinical relevance, saving time and resources compared to the rest of the scales.