Reproducibility, and sensitivity to motor unit loss in amyotrophic lateral sclerosis, of a novel MUNE method: MScanFit MUNE

To examine inter- and intra-rater reproducibility and sensitivity to motor unit loss of a novel motor unit number estimation (MUNE) method, MScanFit MUNE (MScan), compared to two traditional MUNE methods; Multiple point stimulation MUNE (MPS) and Motor Unit Number Index (MUNIX). Twenty-two ALS patients and 20 sex- and age-matched healthy controls were included. MPS, MUNIX, and MScan were performed twice each by two blinded physicians. Reproducibility of MUNE values was assessed by coefficient of variation (CV) and intra class correlation coefficient (ICC). Ability to detect motor unit loss was assessed by ROC curves and area under the curve (AUC). The times taken for each of the methods were recorded. MScan was more reproducible than MPS and MUNIX both between and within operators. The mean CV for MScan (12.3%) was significantly lower than for MPS (24.7%) or MUNIX (21.5%). All methods had ICC > 0.94. MScan and Munix were significantly quicker to perform than MPS (6.3 m vs. 13.2 m). MScan (AUC = 0.930) and MPS (AUC = 0.899) were significantly better at discriminating between patients and healthy controls than MUNIX (AUC = 0.831). MScan was more consistent than MPS or MUNIX and better at distinguishing ALS patients from healthy subjects. MScan may improve detection and assessment of motor unit loss.