Structural and Metabolic Pattern Classification for Detection of Glioblastoma Recurrence and Treatment-Related Effects
- Authors: Jovanovic M.1, Selmic M.2, Macura D.2, Lavrnic S.1, Gavrilovic S.1, Dakovic M.3, Radenkovic S.4, Soldatovic I.5,6, Stosic-Opincal T.6, Maksimovic R.1,6
-
Affiliations:
- Clinical Centre of Serbia, MRI Centre
- Faculty of Transport and Traffic Engineering, University of Belgrade
- Faculty of Physical Chemistry, University of Belgrade
- Institute of Oncology and Radiology of Serbia, Department of Radiation Oncology and Diagnostics
- Institute for Medical Statistics and Informatics
- Medical Faculty, University of Belgrade
- Issue: Vol 48, No 9 (2017)
- Pages: 921-931
- Section: Original Paper
- URL: https://journals.rcsi.science/0937-9347/article/view/247836
- DOI: https://doi.org/10.1007/s00723-017-0913-x
- ID: 247836
Cite item
Abstract
Artificial neuronal network (ANN) in classification of glioblastoma multiforme (GBM) recurrence from treatment effects using advanced magnetic resonance imaging techniques was evaluated. In 56 patients with treated GBM, normalised minimal and mean apparent-diffusion coefficient (ADC) values, vessels number on susceptibility-weighted images (SWI) and Cho/Cr ratio were analysed statistically and by ANN. Significant correlation exists between normalised minimal and mean ADC values, and no correlation between ADC and Cho/Cr values. Cut-off values for tumour presence were: 1.14 for normalised minimal ADC (54% sensitivity, 71% specificity), 1.13 for normalised mean ADC (51% sensitivity, 71% specificity), 1.8 for Cho/Cr ratio (92% sensitivity, 82% specificity), grade 2 for SWI (87% sensitivity, 82% specificity). An accurate prediction of ANN to classify patients into GBM progression or treatment effects group was 99% during the training and 96.8% during the testing phase. Multi-parametric ANN allows distinction between GBM recurrence and treatment effects, and can be used in clinical practice.
About the authors
Marija Jovanovic
Clinical Centre of Serbia, MRI Centre
Author for correspondence.
Email: macvanskimarija@yahoo.com
ORCID iD: 0000-0003-3014-6775
Serbia, Pasterova 2, Belgrade
Milica Selmic
Faculty of Transport and Traffic Engineering, University of Belgrade
Email: macvanskimarija@yahoo.com
Serbia, Vojvode Stepe 305, Belgrade
Dragana Macura
Faculty of Transport and Traffic Engineering, University of Belgrade
Email: macvanskimarija@yahoo.com
Serbia, Vojvode Stepe 305, Belgrade
Slobodan Lavrnic
Clinical Centre of Serbia, MRI Centre
Email: macvanskimarija@yahoo.com
Serbia, Pasterova 2, Belgrade
Svetlana Gavrilovic
Clinical Centre of Serbia, MRI Centre
Email: macvanskimarija@yahoo.com
Serbia, Pasterova 2, Belgrade
Marko Dakovic
Faculty of Physical Chemistry, University of Belgrade
Email: macvanskimarija@yahoo.com
Serbia, Studentski trg 12-16, Belgrade
Sandra Radenkovic
Institute of Oncology and Radiology of Serbia, Department of Radiation Oncology and Diagnostics
Email: macvanskimarija@yahoo.com
Serbia, Pasterova 14, Belgrade
Ivan Soldatovic
Institute for Medical Statistics and Informatics; Medical Faculty, University of Belgrade
Email: macvanskimarija@yahoo.com
Serbia, Dr Subotica 15, Belgrade; Dr Subotica 8, Belgrade
Tatjana Stosic-Opincal
Medical Faculty, University of Belgrade
Email: macvanskimarija@yahoo.com
Serbia, Dr Subotica 8, Belgrade
Ruzica Maksimovic
Clinical Centre of Serbia, MRI Centre; Medical Faculty, University of Belgrade
Email: macvanskimarija@yahoo.com
Serbia, Pasterova 2, Belgrade; Dr Subotica 8, Belgrade