Magnetic resonance imaging and transvaginal ultrasound for determining fibroid burden: implications for research and clinical care.

Abstract

To compare magnetic resonance and ultrasound imaging for uterine fibroid measurement. Eighteen women undergoing hysterectomy for symptomatic fibroids underwent preoperative pelvic ultrasound and magnetic resonance imaging. Resected fibroids were correlated with the images. Weighted kappa agreement statistics and Spearman correlations for patient characteristics were calculated. Magnetic resonance imaging identified 121 of 151 pathologically confirmed fibroids, yielding 91% positive predictive value (95% confidence interval [CI], 85-95) and 80% sensitivity (95% CI, 73-86). Positive predictive value and sensitivity for ultrasound were 97% (95% CI, 89-100) and 40% (95% CI, 32-48), respectively. Mean diameter-equivalent discrepancies between imaging and pathologic measurements were 0.51 +/- 0.68 cm for magnetic resonance imaging and 0.76 +/- 0.88 cm for ultrasound. kappa statistics comparing imaging to pathology showed better agreement for magnetic resonance than ultrasound (kappa = 0.60 vs 0.36). The number of fibroids detected by magnetic resonance imaging predicted measurement errors (r = 0.76; P = .0002). Superior sensitivity and minimal measurement discrepancies suggest magnetic resonance imaging may be preferentially used for fibroid assessment in clinical research.