One of the downsides of diagnostic imaging is that false positives result in additional exams being conducted, driving up the cost of healthcare and raising the stress level of already anxious patients. In this light, interpreting physicians diligently work to raise their awareness of procedures that may cause false positives. Clinicians in Germany have discovered that the advanced imaging technique of time-of flight (TOA) cranial MR angiography (MRA) holds a diagnostic pitfall for primary T1 hyperintense lesions adjacent to a vessel, according to a recent article published in the American Journal of Roentgenology.
“Small intracranial lipomas close to a cerebral artery are hyperintense on TOF MR images and could be mistaken for partially thrombosed aneurysms and associated flow-related artifact,” the authors wrote.
Researchers from the department of neuroradiology at University Hospital Mannheim at the University of Heidelberg in Mannheim, Germany, systematically analyzed the characteristics of intracranial lipomas on TOF images.
“MRA and TOF angiography, in particular, are prone to artifacts that significantly limit reliable identification of aneurysms smaller than 3 mm,” they wrote. “Diagnostic pitfalls during evaluation of 3D TOF MRA may arise from phenomena involving high-signal structures mimicking vascular abnormalities on TOF source images or maximum- intensity-projection (MIP) reconstructions. Possible phenomena include subacute thrombus, complex flow artifacts, and high-signal structures such as lipomas.”
The group conducted a retrospective analysis of 38,000 cranial MRI scans from January 2000- January 2007 and found 17 intracranial lipomas. In addition, scans were conducted on three patients referred to the facility on the basis of suspicious findings on MRA. Images were analyzed by two independent neuroradiologists experienced in interpreting cranial MRI exams.
“The following inclusion criteria were applied to confirm the reported diagnosis of lipoma: first, high-signal intensity of the lesion on T1-weighted images matching the signal intensity of subcutaneous fat and T2-weighted isointensity to fat; and, second signal loss of the lesion on additional T1-weighted fat-suppressed images or hypodensity of –20 H or less on CT images,” they wrote.
They found that TOF source images and MIP reconstructions of the three patients referred to the department showed small spherical lesions adjacent to a vertebrobasilar vessel.
“All lesions exhibited a continuous rim of very low signal, whereas the center had a high signal comparable to lumen of perfused or thrombosed vessel,” they wrote.
The researchers recommended that clinicians consider conducting additional cross-sectional imaging with CT if presented with these indications on cranial TOF MRA.
“Although cross-sectional imaging of lipomas is generally very characteristic, lesions smaller than 5 mm localized close to a vertebrobasilar artery may be misinterpreted as aneurysms on TOF MRA if additional T1 fat-suppressed series or CT images are not available for definite diagnosis,” they suggested.