Evaluation of liver tumor identification rate of volumetric-interpolated breath-hold images using the compressed sensing method and qualitative evaluation of tumor contrast effect via visual evaluation.

Abstract

To evaluate the possible clinical use of the compressed sensing-volumetric-interpolated breath-hold examination (CS-VIBE) in patients with liver tumors by evaluating tumor contrast enhancement effect by radiologists. We examined 22 patients with pathologically confirmed neoplastic lesions in the liver and 62 patients with lesions confirmed by imaging and clinical observation. To evaluate image quality, dynamic contrast-enhanced multiphase breath-hold magnetic resonance imaging was performed. The contrast agent used in this study was gadolinium-ethoxybenzyl-diethylenetriamine penta-acetic acid. Image quality was assessed by three radiologists experienced in this field. Using a four-point scale, we evaluated the gradual contrast enhancement effect of the portal vein to determine whether a decent arterial phase could be obtained. We assessed interobserver agreement using the Fleiss kappa to evaluate image quality between readers. The detection and evaluation of the tumor itself by its enhancement effects are very important in contrast studies. Thus, we evaluated the contrast enhancement effect of the tumors on a three-point scale in 26 patients already known to have hypervascularized tumors using ultrasound or computed tomography as assessed by experienced radiologists. In terms of contrast enhancement effects of the portal vein, the mean value of the readers was 1.85 in the first phase, 2.07 in the second phase, 2.66 in the third phase, 3.05 in the fourth phase, and 3.24 in the fifth phase. Moreover, the interreader agreement was moderate (kappa 0.400-0.502) for all evaluated scores. In the signal of the portal vein, the score of the second arterial phase increased gradually, and in the third arterial phase, the mean score varied from 2 to 3. Compared with ultrasound or computed tomography, CS-VIBE identified 92.3% tumors with hypervascularized tumors (24 of 26 patients with findings hypervascularized tumors). In the results, the interreader agreement was fair to moderate (kappa 0.414-0.521). We obtained multiphase images, including at least one phase, which are useful for the evaluation of liver tumors. Furthermore, the radiologist was able to detect the tumor as before. Therefore, compressed sensing-volumetric-interpolated breath-hold examination is clinically useful in Ethoxybenzyl liver studies.