Publications
2019
1.
Ortiz-Ramón, Rafael; del Carmen Valdés-Hernández, Maria; González-Castro, Victor; Makin, Stephen; Armitage, Paul A; Aribisala, Benjamin S; Bastin, Mark E; Deary, Ian J; Wardlaw, Joanna M; Moratal, David
Identification of the presence of ischaemic stroke lesions by means of texture analysis on brain magnetic resonance images Artículo de revista
En: Computerized Medical Imaging and Graphics, vol. 74, pp. 12–24, 2019, (Publisher: Pergamon).
Resumen | Enlaces | BibTeX | Etiquetas: Radiomics, small vessel disease, Stroke, texture analysis, white matter hyperintensities
@article{ortiz-ramon_identification_2019,
title = {Identification of the presence of ischaemic stroke lesions by means of texture analysis on brain magnetic resonance images},
author = {Rafael Ortiz-Ramón and Maria del Carmen Valdés-Hernández and Victor González-Castro and Stephen Makin and Paul A Armitage and Benjamin S Aribisala and Mark E Bastin and Ian J Deary and Joanna M Wardlaw and David Moratal},
url = {https://www.sciencedirect.com/science/article/pii/S0895611119300278},
year = {2019},
date = {2019-01-01},
journal = {Computerized Medical Imaging and Graphics},
volume = {74},
pages = {12–24},
abstract = {This study investigates using radiomics to detect stroke lesions in brain MRI scans, which are often missed by automated methods. Analyzing 1800 MRI scans, the research found that radiomic features could identify stroke lesions with accuracy between 0.7 and 0.83 AUC. Age was the clinical factor most correlated with accurate detection. The study suggests that incorporating texture features into deep learning models could improve stroke lesion detection in MRI scans.},
note = {Publisher: Pergamon},
keywords = {Radiomics, small vessel disease, Stroke, texture analysis, white matter hyperintensities},
pubstate = {published},
tppubtype = {article}
}
This study investigates using radiomics to detect stroke lesions in brain MRI scans, which are often missed by automated methods. Analyzing 1800 MRI scans, the research found that radiomic features could identify stroke lesions with accuracy between 0.7 and 0.83 AUC. Age was the clinical factor most correlated with accurate detection. The study suggests that incorporating texture features into deep learning models could improve stroke lesion detection in MRI scans.