Flavio Caprioli MD, Abstract Co-Author: Nothing to Disclose

Alessandra Losco MD, Abstract Co-Author: Nothing to Disclose

Guido Basilisco MD, Abstract Co-Author: Nothing to Disclose

Chiara Viganò MD, Abstract Co-Author: Nothing to Disclose

Pietro Raimondo Biondetti MD, Presenter: Nothing to Disclose

Laura Virginia Forzenigo MD, Abstract Co-Author: Nothing to Disclose

Dario Conte MD, Abstract Co-Author: Nothing to Disclose

et al, Abstract Co-Author: Nothing to Disclose

To assess, in a consecutive series of patients with perianal fistulizing Crohn’s disease (CD): 1) the diagnostic performance of a visual assessment of fistula track activity at anal ultrasound (AUS) in comparison to magnetic resonance imaging (MRI) T2-hyperintensity. 2) whether the application of a quantitative computerized analysis of grey-scale tones (GST) might improve the visual assessment.

28 consecutive patients (mean age 37 years, range 20-62) with a CD-related perianal fistula were studied. Patients underwent MRI (Magnetom SP 63 1.5 Tesla, Siemens Avanto©, Germany) with T2-weighted axial images acquisition as reference standard for fistula activity. AUS was performed with Falcon 2101 EXL equipped with a 1850 10-Mhz rotating transducer (Brüel and Kjær©, Herlev, Denmark). Visual assessment of fistula track activity at AUS was performed by two independent operators. Regions of interest (ROI) including the area of fistula tracks were selected on AUS images and mean grey-scale tones (GST) of the selected ROIs were calculated for each patient with a digitized image analysis software.

Sensitivity values for the visual assessment in the identification of active perianal fistulas were 0.85 – 0.90 and specificity 0.62-0.75; calculated likelihood ratios for positive and negative test results were 2.4-3.4 and 0.16-0.20 for the two observers, respectively. At the computer-assisted analysis of GST, hyperintense fistulas were significantly more hypoechoic than hypointense ones (P<0.001). The computer-assisted analysis sensitivity was 0.85-0.95 and specificity 0.88 for both operators; likelihood ratios for positive and negative test results were 7.1-7.9 and 0.06-0.17. Interobserver agreement for the visual assessment of fistula activity between the two operators was good (k value=0.652).

Visual assessment of AUS images is a sensitive method to evaluate fistula track activity in patients with CD. The application of a computer-assisted evaluation of GST confirms the results of visual assessment and improves the specificity of the evaluation.

Caprioli, F, Losco, A, Basilisco, G, Viganò, C, Biondetti, P, Forzenigo, L, Conte, D, et al, , Computer-assisted Evaluation of Perianal Fistula Activity with Anal Ultrasound in Patients with Crohn’s Disease.  Radiological Society of North America 2005 Scientific Assembly and Annual Meeting, November 27 - December 2, 2005 ,Chicago IL. http://archive.rsna.org/2005/4405656.html