Premium
Extranodal spread in the neck: MRI detection on the basis of pixel‐based time‐signal intensity curve analysis
Author(s) -
Sumi Misa,
Nakamura Takashi
Publication year - 2011
Publication title -
journal of magnetic resonance imaging
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.563
H-Index - 160
eISSN - 1522-2586
pISSN - 1053-1807
DOI - 10.1002/jmri.22454
Subject(s) - magnetic resonance imaging , medicine , washout , nuclear medicine , dynamic contrast , logistic regression , radiology
Purpose We evaluated dynamic contrast‐enhanced (DCE) magnetic resonance imaging (MRI) for the preoperative detection of extranodal spread (ENS) in metastatic nodes in the neck. Materials and Methods The time‐signal intensity curve (TIC) profiles of 54 histologically proven metastatic nodes (26 ENS‐positive and 28 ENS‐negative) from 43 patients with head and neck squamous cell carcinoma (SCC) were retrospectively analyzed to determine the effective TIC criteria for ENS‐positive nodes. The TICs were semiautomatically classified into four distinctive patterns (flat, slow uptake, rapid uptake with low washout ratio, and rapid uptake with high washout ratio) on a pixel‐by‐pixel basis. Results A number of the MRI findings were significantly correlated with ENS. However, multivariate logistic regression analysis revealed that only a short‐axis diameter and an area with slow uptake TIC patterns were significantly and independently indicative of the presence of ENS. The combined MRI criteria of nodal size (>25 mm) or TIC profile (>44% nodal areas with slow‐uptake TIC patterns) yielded the best results for differentiation between ENS‐positive and ENS‐negative nodes, providing 96% sensitivity, 100% specificity, 98% accuracy, and 100% positive, and 97% negative predictive values. Conclusion When combined with size criteria, pixel‐based MR factor analysis may be a promising tool for detecting ENS. J. Magn. Reson. Imaging 2011;33:830–838. © 2011 Wiley‐Liss, Inc.