Impact of Bent Distortion on Accuracy of Measurement During Transrectal Ultrasonography for Prostatic Imaging: A Preliminary Study

Abstract

OBJECTIVE: To evaluate the effect of bent distortion of the image of the prostate on the accuracy of 3-dimensional measurements during transrectal ultrasonography (TRUS). METHODS: TRUS images were obtained prospectively from 60 patients with prostate cancer. The degree of rectal compression on the probe was varied in each case to obtain a typical bent distortion image (BDI) and a minimally bent image (MBI). Magnetic resonance imaging (MRI) served as the reference. Axial and midsagittal BDI, MDI, and MRI were selected. Three-dimensional prostate diameters (height, width, and length) and volume were obtained from all images by 2 independent observers. The data were analyzed by using a paired t test, intraclass correlation coefficients (ICCs), and Bland-Altman plots. RESULTS: The MBI and BDI diameters differed significantly (P <.001). All ICCs for the MBI diameters and volume and the ICCs for BDI length and volume exceeded 0.9, which indicates high measurement reliability. However, the ICCs for BDI height and width were below 0.9. When the difference in MBI or BDI relative to MRI was calculated for each variable, the MBI differences were all significantly smaller than the BDI differences (P <.001). CONCLUSION: Bent distortion during TRUS affects the accuracy of measurements. MBI is also morphologically more realistic than BDI. Therefore, to minimize prostatic image distortion, the rectal probe should be placed so that the posterior wall of the prostate is as flat as possible.