ABSTRACT
PURPOSE: To evaluate the added value of diffusion-weighted (DWI) magnetic resonance imaging (MRI) by comparison with T2-weighted images alone in the diagnosis of perianal fistula. MATERIAL AND METHODS: MRI examinations of 123 patients (97 men, 26 women; mean age, 41.9 years) with suspected perianal fistula were retrospectively evaluated by two radiologists. Fat-suppressed T2-weighted fast spin echo images, DWI (b values, 0 and 1000s/mm2) and fat-suppressed gadolinium chelate-enhanced T1-weighted images were evaluated for each patient by using a four-point scale. Confidence scores and sensitivities were calculated for T2-weighted images alone, the combination of DWI and T2-weighted images and the combination of gadolinium chelate-enhanced T1-weighted images and T2-weighted images. The combination of gadolinium chelate-enhanced and T2-weighted images was used as reference standard. RESULTS: Perianal fistulas were present in 92/123 patients (74.8%). An almost perfect interobserver agreement was found for T2-weighted images (kappa=0.868), the combination of gadolinium chelate-enhanced T1-weighted images and T2-weighted images (kappa=0.96) and the combination of DWI and T2-weighted images (kappa=0.90). The confidence scores for the diagnosis of perianal fistula for the combination of gadolinium chelate-enhanced T1-weighted images and T2-weighted images were greater than those of T2-weighted images alone for observer 1 (P<0.001) and observer 2 (P=0.009). The confidence scores of the combination of DWI and T2-weighted images were greater than those of T2-weighted images alone for observer 1 (P<0.001) and observer 2 (P=0.032). Sensitivity and specificity of the combination of DWI and T2-weighted images were greater than those of T2-weighted images alone for both observers. CONCLUSION: DWI has a significant added value compared to T2-weighted imaging alone in the diagnosis of perianal fistula.
