Magnetic resonance imaging subtraction vs. pre- and post-contrast 3D gradient recalled echo fat suppressed imaging for evaluation of the canine and feline brain

Abstract

Subtraction magnetic resonance imaging (MRI) has been reported to increase accuracy in the diagnosis of meningeal and inflammatory brain diseases in small animals. 3D T1W gradient recalled echo (GRE) techniques have been proposed as a suitable alternative to conventional spin echo sequences in imaging the canine brain. The aim of this study was to compare subtraction images and paired pre- and post-contrast 3D T1W GRE fat suppressed (FS) images in canine and feline MRI studies using clinical diagnosis as the gold standard. Paired pre- and post-contrast T1W 3D FS GRE images and individual subtraction images of 100 small animal patients were randomized and independently evaluated by 2 blinded observers. Diagnosis categories were “normal,” “inflammatory,” “neoplastic,” and “other.” Clinical diagnosis was made in the same categories and served as the gold standard. Image interpretation results were compared to the clinical diagnosis. Interobserver agreement was determined. Clinically, 41 studies were categorized as “normal,” 18 as “inflammatory,” 28 as “neoplastic,” and 13 as “other.” The agreement of the pre- and post-contrast GRE images with the gold standard was significantly higher than that of the subtraction images (k = 0.7491 vs. k = 0.5924; p = 0.0075). The largest sources of error were misinterpretation of “other” as “normal” and “normal” as “inflammatory.” There was no significant difference between the two observers (p = 0.8820). Based on this study, subtraction images do not provide an advantage to paired pre- and post-contrast FS GRE images when evaluating the canine and feline brain.