To retrospectively correlate the magnetic resonance imaging (MRI) diagnosis with the surgical findings and/or clinical outcomes in patients presenting with an acute scrotum.

From January 1997 to June 2004, 39 selected patients presenting with an acute scrotum underwent dynamic contrast-enhanced subtraction MRI as a 24-hour stand-by examination. The MRI diagnosis was based solely on the presence or absence of testicular contrast enhancement, without respect to the clinical history or physical examination findings.

Contrast enhancement of the affected and normal testes could be accurately compared in all cases. The MRI diagnosis was testicular torsion (no testicular contrast enhancement, n = 13), nonspecific (reduced to normal enhancement, n = 10), appendiceal torsion (n = 3), and epididymitis (n = 13). The surgical findings and/or subsequent clinical outcomes showed that MRI diagnosis of testicular torsion was accurate in all 13 cases. Furthermore, exploration revealed testicular torsion in 1 case with very little enhancement. The sensitivity and specificity of MRI in the diagnosis of testicular torsion was 93% (13 of 14) and 100% (25 of 25), respectively. Although MRI detected testicular perfusion, 5 of the 6 patients with clinical suspicion of intermittent torsion underwent surgical exploration.

MRI is a highly accurate imaging modality for the diagnosis of testicular torsion. However, it could not be used to rule out intermittent torsion and the clinical use of negative findings for an MRI torsion test was less than 100% specificity. Because this was a preliminary retrospective study, the true clinical value of MRI should be evaluated by prospective studies.