Paolo Florent Felisaz1, Francesco Balducci2, Salvatore Gitto2, Irene Carne3, Stefano Montagna3, Roberto De Icco2, Anna Pichiecchio4, Maurizia Baldi3, Fabrizio Calliada5, Stefano Bastianello6. 1. Institute of Radiology, University of Pavia, Viale Camillo Golgi 19, 27100 Pavia PV, Italy. Electronic address: paolo.felisaz@gmail.com. 2. Institute of Radiology, University of Pavia, Viale Camillo Golgi 19, 27100 Pavia PV, Italy. 3. Department of Radiology, IRCCS Salvatore Maugeri Foundation, Scientific Institute of Pavia, Pavia PV, Italy. 4. Department of Neuroradiology, C. Mondino National Neurological Institute, Pavia PV, Italy. 5. Institute of Radiology, University of Pavia, Viale Camillo Golgi 19, 27100 Pavia PV, Italy; Institute of Radiology, IRCCS Foundation Policlinico San Matteo, Pavia PV, Italy. 6. Department of Neuroradiology, C. Mondino National Neurological Institute, Pavia PV, Italy; Department of Brain and Behavioral Sciences, University of Pavia, Pavia PV, Italy.
Abstract
RATIONALE AND OBJECTIVES: The aims of this study were to propose a semiautomated technique to segment and measure the volume of different nerve components of the tibial nerve, such as the nerve fascicles and the epineurium, based on magnetic resonance microneurography and a segmentation tool derived from brain imaging; and to assess the reliability of this method by measuring interobserver and intraobserver agreement. MATERIALS AND METHODS: The tibial nerve of 20 healthy volunteers (age range = 23-69; mean = 47; standard deviation = 15) was investigated at the ankle level. High-resolution images were obtained through tailored microneurographic sequences, covering 28 mm of nerve length. Two operators manually segmented the nerve using the in-phase image. This region of interest was used to mask the nerve in the water image, and two-class segmentation was performed to measure the fascicular volume, epineurial volume, nerve volume, and fascicular to nerve volume ratio (FNR). Interobserver and intraobserver agreements were calculated. RESULTS: The nerve structure was clearly visualized with distinction of the fascicles and the epineurium. Segmentation provided absolute volumes for nerve volume, fascicular volume, and epineurial volume. The mean FNR resulted in 0.69 with a standard deviation of 0.04 and appeared to be not correlated with age and sex. Interobserver and intraobserver agreements were excellent with alpha values >0.9 for each parameter investigated, with measurements free of systematic errors at the Bland-Altman analysis. CONCLUSIONS: We concluded that the method is reproducible and the parameter FNR is a novel feature that may help in the diagnosis of neuropathies detecting changes in volume of the fascicles or the epineurium.
RATIONALE AND OBJECTIVES: The aims of this study were to propose a semiautomated technique to segment and measure the volume of different nerve components of the tibial nerve, such as the nerve fascicles and the epineurium, based on magnetic resonance microneurography and a segmentation tool derived from brain imaging; and to assess the reliability of this method by measuring interobserver and intraobserver agreement. MATERIALS AND METHODS: The tibial nerve of 20 healthy volunteers (age range = 23-69; mean = 47; standard deviation = 15) was investigated at the ankle level. High-resolution images were obtained through tailored microneurographic sequences, covering 28 mm of nerve length. Two operators manually segmented the nerve using the in-phase image. This region of interest was used to mask the nerve in the water image, and two-class segmentation was performed to measure the fascicular volume, epineurial volume, nerve volume, and fascicular to nerve volume ratio (FNR). Interobserver and intraobserver agreements were calculated. RESULTS: The nerve structure was clearly visualized with distinction of the fascicles and the epineurium. Segmentation provided absolute volumes for nerve volume, fascicular volume, and epineurial volume. The mean FNR resulted in 0.69 with a standard deviation of 0.04 and appeared to be not correlated with age and sex. Interobserver and intraobserver agreements were excellent with alpha values >0.9 for each parameter investigated, with measurements free of systematic errors at the Bland-Altman analysis. CONCLUSIONS: We concluded that the method is reproducible and the parameter FNR is a novel feature that may help in the diagnosis of neuropathies detecting changes in volume of the fascicles or the epineurium.
Authors: P F Felisaz; G Maugeri; V Busi; R Vitale; F Balducci; S Gitto; P Leporati; A Pichiecchio; M Baldi; F Calliada; L Chiovato; S Bastianello Journal: Radiol Res Pract Date: 2017-04-16