A combined MRI biomarker approach using a non-standard Multiple Factor Analysis

Abstract : In medical applications, various MRI biomarkers, which are extracted from different MRI modalities, are used to detect physiologic abnormalities but different biomarkers are usually sensitive to different aspects of the disease. Thus, it would be interesting to use the information carried by multiple biomarkers, especially in the context of clinical trials where the efficiency of a treatment should be judged as precisely as possible. This paper proposes an approach that combined various MRI biomarkers in the context of a Multiple Sclerosis (MS) clinical trial. The method mainly includes four steps: After extracting the biomarkers from different MRI modalities, a histogram analysis is performed followed by a Multiple Factor Analysis (MFA) to produce linear combinations of the MRI biomarkers and finally, a Hierarchical Clustering based on the MFA results is executed. The aim of this approach is to conclude more effectively on the effect of a treatment in a clinical trial.
Document type :
Conference papers
Liste complète des métadonnées

Contributor : Laurence Porte <>
Submitted on : Tuesday, January 29, 2019 - 2:04:25 PM
Last modification on : Friday, April 12, 2019 - 4:22:52 PM


Files produced by the author(s)



Sana Rebbah, Daniel Delahaye, Stéphane Puechmorel, Florence Nicol, Pierre Maréchal, et al.. A combined MRI biomarker approach using a non-standard Multiple Factor Analysis. CISP-BMEI 2018, 11th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics, Oct 2018, Beijing, China. pp.art. no. 8633261., ⟨10.1109/CISP-BMEI.2018.8633261⟩. ⟨hal-01998086⟩



Record views


Files downloads