Valeria Elisa Contarino1, Giorgio Conte2, Claudia Morelli3, Francesca Trogu3, Elisa Scola1, Sonia Francesca Calloni4, Luis Carlos Sanmiguel Serpa5, Chunlei Liu6, Vincenzo Silani7, Fabio Triulzi8. 1. Fondazione IRCCS Ca' Granda Ospedale Maggiore Policlinico, Neuroradiology Unit, Via Francesco Sforza 35, Milan, Italy. 2. Fondazione IRCCS Ca' Granda Ospedale Maggiore Policlinico, Neuroradiology Unit, Via Francesco Sforza 35, Milan, Italy. Electronic address: giorgioconte.unimed@gmail.com. 3. IRCCS Istituto Auxologico Italiano, Ospedale San Luca, Neurology Unit, Piazziale Brescia 20, Milan, Italy. 4. San Raffaele Scientific Institute, Department of Neuroradiology, Via Olgettina 60, Milan, Italy. 5. Department of Electronics, Information and Bioengineering, Politecnico Di Milano, Via Giuseppe Ponzio 34, Milan, Italy. 6. Department of Electrical Engineering and Computer Sciences, University of California, Berkeley, CA, 94720, USA. 7. IRCCS Istituto Auxologico Italiano, Ospedale San Luca, Neurology Unit, Piazziale Brescia 20, Milan, Italy; Department of Pathophysiology and Transplantation, Università Degli Studi Di Milano, Italy. 8. Fondazione IRCCS Ca' Granda Ospedale Maggiore Policlinico, Neuroradiology Unit, Via Francesco Sforza 35, Milan, Italy; Department of Pathophysiology and Transplantation, Università Degli Studi Di Milano, Italy.
Abstract
PURPOSE: Diagnostic work-up in motor neuron disease (MND) needs a quantitative biomarker of upper motor neuron (UMN) impairment. We investigated the susceptibility properties of the precentral cortex in a cohort of patients affected by Amyotrophic lateral sclerosis (ALS) to obtain a useful biomarker of UMN impairment in a fully automatic paradigm. MATERIALS AND METHODS: We retrospectively collected imaging and clinical data of 42 ALS patients who had undergone brain 3 T MRI including tridimensional T1-weighted and spoiled gradient-echo multi-echo T2-weighted images. We further acquired images from 23 healthy control (HC) volunteers. The precentral cortex was automatically segmented and the cortical thickness calculated. Histogram metrics (mean, median, standard deviation, skewness, kurtosis) derived from the quantitative susceptibility map (QSM) were extracted from the automatically segmented precentral cortex. Multivariate regression analyses were performed to identify the variables predicting the disease status (ALS vs HC), the Amyotrophic Lateral Sclerosis Functional Rating Scale-Revised (ALSFRS-R) and the UMN score. RESULTS: A decreased cortical thickness (B = 9.40; Wald's test = 7.43; p = 0.006) and increased susceptibility skewness (B = -3.08; Wald's test = 4.36; p = 0.037) independently predicted ALS in a logistic regression model (χ2(3, N = 65) = 22.07, p < 0.001. No predictors of ALSFRS-R were identified. An increased susceptibility skewness (β = 0.55; t = 4.23; p < 0.001) and longer disease duration (β = 0.35; t = 2.67; p = 0.011) independently predicted a higher UMN score in a linear regression model (R2 = 0.32; p < 0.001). CONCLUSION: The susceptibility skewness might be an unbiased quantitative biomarker of UMN impairment in ALS patients.
PURPOSE: Diagnostic work-up in motor neuron disease (MND) needs a quantitative biomarker of upper motor neuron (UMN) impairment. We investigated the susceptibility properties of the precentral cortex in a cohort of patients affected by Amyotrophic lateral sclerosis (ALS) to obtain a useful biomarker of UMN impairment in a fully automatic paradigm. MATERIALS AND METHODS: We retrospectively collected imaging and clinical data of 42 ALSpatients who had undergone brain 3 T MRI including tridimensional T1-weighted and spoiled gradient-echo multi-echo T2-weighted images. We further acquired images from 23 healthy control (HC) volunteers. The precentral cortex was automatically segmented and the cortical thickness calculated. Histogram metrics (mean, median, standard deviation, skewness, kurtosis) derived from the quantitative susceptibility map (QSM) were extracted from the automatically segmented precentral cortex. Multivariate regression analyses were performed to identify the variables predicting the disease status (ALS vs HC), the Amyotrophic Lateral Sclerosis Functional Rating Scale-Revised (ALSFRS-R) and the UMN score. RESULTS: A decreased cortical thickness (B = 9.40; Wald's test = 7.43; p = 0.006) and increased susceptibility skewness (B = -3.08; Wald's test = 4.36; p = 0.037) independently predicted ALS in a logistic regression model (χ2(3, N = 65) = 22.07, p < 0.001. No predictors of ALSFRS-R were identified. An increased susceptibility skewness (β = 0.55; t = 4.23; p < 0.001) and longer disease duration (β = 0.35; t = 2.67; p = 0.011) independently predicted a higher UMN score in a linear regression model (R2 = 0.32; p < 0.001). CONCLUSION: The susceptibility skewness might be an unbiased quantitative biomarker of UMN impairment in ALSpatients.