Cargando…

A Magnetic Resonance-Relaxometry-Based Technique to Identify Blood Products in Brain Parenchyma: An Experimental Study on a Rabbit Model

Magnetic resonance relaxometry is a quantitative technique that estimates T1/T2 tissue relaxation times. This has been proven to increase MRI diagnostic accuracy of brain disorders in human medicine. However, literature in the veterinary field is scarce. In this work, a T1 and T2-based relaxometry a...

Descripción completa

Detalles Bibliográficos
Autores principales: Del Signore, Francesca, Vignoli, Massimo, Della Salda, Leonardo, Tamburro, Roberto, Paolini, Andrea, Cerasoli, Ilaria, Chincarini, Matteo, Rossi, Emanuela, Ferri, Nicola, Romanucci, Mariarita, Falerno, Ilaria, de Pasquale, Francesco
Formato: Online Artículo Texto
Lenguaje:English
Publicado: Frontiers Media S.A. 2022
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9195168/
https://www.ncbi.nlm.nih.gov/pubmed/35711807
http://dx.doi.org/10.3389/fvets.2022.802272
Descripción
Sumario:Magnetic resonance relaxometry is a quantitative technique that estimates T1/T2 tissue relaxation times. This has been proven to increase MRI diagnostic accuracy of brain disorders in human medicine. However, literature in the veterinary field is scarce. In this work, a T1 and T2-based relaxometry approach has been developed. The aim is to investigate its performance in characterizing subtle brain lesions obtained with autologous blood injections in rabbits. This study was performed with a low-field scanner, typically present in veterinary clinics. The approach consisted of a semi-automatic hierarchical classification of different regions, selected from a T2 map. The classification was driven according to the relaxometry properties extracted from a set of regions selected by the radiologist to compare the suspected lesion with the healthy parenchyma. Histopathological analyses were performed to estimate the performance of the proposed classifier through receiver operating characteristic curve analyses. The classifier resulted in moderate accuracy in terms of lesion characterization.