Quantitative neuroimaging in multiple sclerosis (MS): which biomarker is the most reliable in the evaluation of MS-induced atrophy?
Author(s): ,
E. Makras
General Anticancer and Oncological Hospital of Athens `St. Savvas`
M. Pelechrini
MRI and Neuroradiology, General Anticancer and Oncological Hospital of Athens `St. Savvas`
A.-A. Katsarou
Euroclinic of Athens, Athens, Greece
K. Pelechrinis
School of Computing and Information, University of Pittsburgh, Pittsburgh, PA, United States
A. Gouliamos
Iatropolis, Athens, Greece
S. Bisdas
University College of London (UCL), London, United Kingdom
V. Katsaros
MRI and Neuroradiology, University of Athens, Athens, Greece
ECTRIMS Online Library. Makras E. Oct 12, 2018; 229130; P1290
Elefterios Makras
Elefterios Makras
Login now to access Regular content available to all registered users.

You may also access this content "anytime, anywhere" with the Free MULTILEARNING App for iOS and Android
Discussion Forum (0)
Rate & Comment (0)

Abstract: P1290

Type: Poster Sessions

Abstract Category: Therapy - Others

Introduction: Brain atrophy induced by MS is a frequent and well-established entity according to international literature.
Objectives: In the era of quantitative neuroimaging/ artificial intelligence assisted diagnosis and precision/personalized medicine, there is a need for tools that will allow us to reliably evaluate brain atrophy.
Aims: Our aim was to determine which quantitative biomarker extracted from whole brain volumetric assessment could be considered as the most reliable in the evaluation of MS induced brain atrophy.
Methods: Volumetric data from 118 examinations of 81 patients (mean age: 41) with an established MS diagnosis, were retrospectively included from four centers (three with 1.5T MRI, one 3T MRI). The available 3D T1 MPRAGE pre-/ 10 minutes post-Gd and FLAIR sequences were uploaded to IcoBrain(TM) platform for volumetric evaluation. . Logistic regression statistical analysis with cross-validation was then used to evaluate the resulting volume percentiles. We randomly split the data into a training set (80%) that we used to build our model and a test set (20%) that we used for its evaluation. We repeated this process 20 times and for each one of them we built three different logistic regression (logit) models for the probability of the presence of atrophy (defined as the bottom 5th percentile of the whole brain volume). Using the test set in each of the splits we calculated the sensitivity, specificity, AUROC and Brier scores respectively and report the average values.
Results: The best performing model had both white and gray matter volume percentiles as independent variables and showed the following average values: AUROC 0.99, Brier score 0.04, sensitivity 94% and specificity 95%. The rest of the models used only one independent variable (either white matter volume percentile or grey matter volume percentile). The model with the white matter volume percentile as an independent variable showed average AUROC 0.85, Brier score 0.16, sensitivity 78% and specificity 77%, while the one with the gray matter volume percentile showed average AUROC 0.86, Brier score 0.15, sensitivity 77% and specificity 79%.
Conclusion: The combination of white and gray matter volume percentiles seems the most reliable biomarker in the evaluation of MS induced atrophy and holds promise as a prognostic one.
Eleftherios Makras: nothing to disclose
Maria Pelechrini: nothing to disclose
Agapi-Alexandra Katsarou: nothing to disclose
Konstantinos Pelechrinis: nothing to disclose
Sotirios Bisdas: nothing to disclose
Vasileios Katsaros: nothing to disclose
Athanasios Gouliamos: nothing to disclose

Code of conduct/disclaimer available in General Terms & Conditions
Anonymous User Privacy Preferences

Strictly Necessary Cookies (Always Active)

MULTILEARNING platforms and tools hereinafter referred as “MLG SOFTWARE” are provided to you as pure educational platforms/services requiring cookies to operate. In the case of the MLG SOFTWARE, cookies are essential for the Platform to function properly for the provision of education. If these cookies are disabled, a large subset of the functionality provided by the Platform will either be unavailable or cease to work as expected. The MLG SOFTWARE do not capture non-essential activities such as menu items and listings you click on or pages viewed.

Performance Cookies

Performance cookies are used to analyse how visitors use a website in order to provide a better user experience.

Google Analytics is used for user behavior tracking/reporting. Google Analytics works in parallel and independently from MLG’s features. Google Analytics relies on cookies and these cookies can be used by Google to track users across different platforms/services.

Save Settings