BACKGROUND AND OBJECTIVE: The persistence of new enhancing T1 hypointense lesions (acute black holes, ABHs) in multiple sclerosis (MS) cannot be predicted visually at lesion onset. Texture analysis using the polar Stockwell transform (PST) applied to conventional MR images however shows promise in quantifying tissue injury early. The objective of this study was to explore whether ABHs that persist (pABHs) differ from those that are transient (tABHs) using PST texture analysis. METHODS: Fifteen ABHs (8 pABHs; 7 tABHs) from 9 patients were analyzed on 3T images obtained during a clinical trial. Persistence was defined as remaining T1 hypointense 5-8 months later. NAWM regions were examined to control for changes unrelated to ABHs. RESULTS: At first appearance, there was higher coarse texture indicating greater tissue damage in the pABHs than in the tABHs (p<0.01). Both had greater coarse texture than the contralateral and general NAWM (p≤0.01). No difference was identified in normalized signal intensity between pABHs and tABHs and neither demonstrated location preference. While tABHs tended to be smaller than pABHs there was no correlation between lesion size and texture (r=0.44, p>0.05). Furthermore, coarse texture content appeared to predict persistence of individual lesions. CONCLUSIONS: This preliminary study suggests that PST texture could predict persistence of tissue injury based on the severity of structural disorganization within acute lesions. While confirmation of this data is required texture analysis may prove to be a valuable tool to quantify tissue damage and predict recovery in proof-of-concept neuroprotection and repair trials.
BACKGROUND AND OBJECTIVE: The persistence of new enhancing T1 hypointense lesions (acute black holes, ABHs) in multiple sclerosis (MS) cannot be predicted visually at lesion onset. Texture analysis using the polar Stockwell transform (PST) applied to conventional MR images however shows promise in quantifying tissue injury early. The objective of this study was to explore whether ABHs that persist (pABHs) differ from those that are transient (tABHs) using PST texture analysis. METHODS: Fifteen ABHs (8 pABHs; 7 tABHs) from 9 patients were analyzed on 3T images obtained during a clinical trial. Persistence was defined as remaining T1 hypointense 5-8 months later. NAWM regions were examined to control for changes unrelated to ABHs. RESULTS: At first appearance, there was higher coarse texture indicating greater tissue damage in the pABHs than in the tABHs (p<0.01). Both had greater coarse texture than the contralateral and general NAWM (p≤0.01). No difference was identified in normalized signal intensity between pABHs and tABHs and neither demonstrated location preference. While tABHs tended to be smaller than pABHs there was no correlation between lesion size and texture (r=0.44, p>0.05). Furthermore, coarse texture content appeared to predict persistence of individual lesions. CONCLUSIONS: This preliminary study suggests that PST texture could predict persistence of tissue injury based on the severity of structural disorganization within acute lesions. While confirmation of this data is required texture analysis may prove to be a valuable tool to quantify tissue damage and predict recovery in proof-of-concept neuroprotection and repair trials.
Authors: Yunyan Zhang; Luanne M Metz; James N Scott; Jessie Trufyn; Gordon H Fick; Fiona Costello Journal: Neuroimage Clin Date: 2014-01-14 Impact factor: 4.881
Authors: Rajeev Kumar Verma; Johannes Slotboom; Mirjam Rachel Heldner; Mirjam Rahel Heldner; Frauke Kellner-Weldon; Raimund Kottke; Christoph Ozdoba; Christian Weisstanner; Christian Philipp Kamm; Roland Wiest Journal: PLoS One Date: 2013-07-16 Impact factor: 3.240