| Literature DB >> 26306004 |
Joe Fenn1, Randi Drees1, Holger A Volk1, Steven De Decker1.
Abstract
Ischemic myelopathy (IM) and acute noncompressive nucleus pulposus extrusion (ANNPE) are common spinal emergencies in dogs with similar clinical presentations. Magnetic resonance imaging (MRI) criteria for a presumptive antemortem diagnosis have been reported, however inter- and intraobserver agreement for use of these criteria has not been established. The aim of this retrospective, descriptive, cross-sectional study was to describe inter- and intraobserver agreement for using previously published MRI criteria to diagnose presumptive IM and ANNPE in a sample dogs. Dogs with a presumptive diagnosis of IM or ANNPE and available MRI scan data were retrieved from medical record archives during the period of 2009 and 2013. A total of 127 dogs were identified. From this sample, MRI scans for 60 dogs were randomly selected and duplicated for intraobserver analysis, giving a total of 187 anonymized studies that were presented to two blinded assessors (one board-certified veterinary neurologist, one board-certified veterinary radiologist). Assessors were asked to diagnose lesions as IM or ANNPE based on previously published MRI characteristics. Interobserver agreement in diagnosing IM or ANNPE was moderate (Kappa = 0.56) and intraobserver agreement was moderate to good (Assessor 1 Kappa = 0.79, Assessor 2 Kappa = 0.47). Agreement was strongest for detecting presence of lesions overlying a vertebral body (94% of lesions that were diagnosed as IM) or overlying an intervertebral disk (85% of lesions that were diagnosed as ANNPE). Findings indicated that use of previously published MRI criteria yields moderate inter- and moderate to good intraobserver agreement for a presumptive diagnosis of IM or ANNPE in dogs.Entities:
Keywords: MRI; fibrocartilaginous embolism; high velocity low volume disk extrusion; traumatic disk extrusion
Mesh:
Year: 2015 PMID: 26306004 DOI: 10.1111/vru.12289
Source DB: PubMed Journal: Vet Radiol Ultrasound ISSN: 1058-8183 Impact factor: 1.363