DIAGNOSTIC NEURORADIOLOGY T2*-weighted MR angiography substantially increases the detection of hemorrhage in the wall of brain abscess: implications in clinical interpretation Rakesh Kumar Gupta & VaishaliTomar & Rishi Awasthi & Abhishek Yadav & Nuzhat Husain & Vikas Bharadwaj & Bal K. Ojha & Sanjay Behari & Kashi N. Prasad & Ram Kishore Singh Rathore Received: 24 May 2011 /Accepted: 27 August 2011 # Springer-Verlag 2011 Abstract Introduction The purpose of the present study was to identify the true prevalence of hemorrhage in the abscess using T2*-weighted angiography (SWAN) imaging and to study its influence on diffusion tensor imaging (DTI) metrics. Methods Fifteen patients of brain abscess underwent conventional, SWAN, and DT imaging on a 3-T MRI followed by its confirmation with histology. DTI metrics were quantified by region-of-interest analysis on hemor- rhagic and non-hemorrhagic regions of the abscess wall. Prussian blue staining was performed on excised abscess walls to confirm hemorrhage on histology. Results Eleven of 15 patients showed evidence of hemor- rhage on both Prussian blue staining as well as SWAN imaging. Fractional anisotropy (FA) and linear anisotropy (CL) values were significantly higher, while spherical anisotropy was significantly lower in hemorrhagic compared to non-hemorrhagic regions of the abscess wall. Conclusion Hemorrhage in the abscess wall is a common feature and may not always indicate neoplasm. The presence of intracellular iron in addition to concentrically laid collagen fibers may have synergistic effect on FA and CL values in the abscess wall. Inclusion of SWAN to MRI protocol will define the true prevalence of hemorrhage in brain abscess. Keywords T2*-weighted angiography (SWAN) imaging . Diffusion tensor imaging . Brain abscess . Hemorrhage . Iron Introduction Brain abscess is a potentially fatal central nervous system infection which grows as a localized area of cerebritis and develops into a collection of pus surrounded by a well- vascularized capsule within the brain parenchyma [1, 2]. In V . Tomar : R. Awasthi : A. Yadav Department of Radiodiagnosis, Sanjay Gandhi Post Graduate Institute of Medical Sciences, Lucknow, Uttar Pradesh 226014, India K. N. Prasad Department of Microbiology, Sanjay Gandhi Post Graduate Institute of Medical Sciences, Lucknow, Uttar Pradesh, India S. Behari Department of Neurosurgery, Sanjay Gandhi Post Graduate Institute of Medical Sciences, Lucknow, Uttar Pradesh, India V . Bharadwaj : B. K. Ojha Department of Neurosurgery, Chhatrapati Sahuji Maharaj Medical University, Lucknow, Uttar Pradesh, India N. Husain Department of Pathology, Chhatrapati Sahuji Maharaj Medical University, Lucknow, Uttar Pradesh, India R. K. S. Rathore Department of Mathematics and Statistics, Indian Institute of Technology, Kanpur, Uttar Pradesh, India R. K. Gupta (*) MR Section, Department of Radiodiagnosis, Sanjay Gandhi Post Graduate Institute of Medical Sciences, Lucknow, Uttar Pradesh 226014, India e-mail: rgupta@sgpgi.ac.in R. K. Gupta e-mail: rakeshree1@gmail.com Neuroradiology DOI 10.1007/s00234-011-0952-1