Usefulness of contrast-enhanced dual-energy computed tomography for detecting oral, head, and neck cancers

Abstract

Purpose Contrast-enhanced dual-energy computed tomography (DECT) is advantageous in classifying tissues by changing iodine density. However, only few studies have investigated the interobserver variations in image quality and reproducibility by several observers (> 10). Therefore, this study aimed to compare iodine density reconstructed computed tomography (IDCT) and non-reconstructed computed tomography (NRCT) in terms of visual scoring and tumor size estimation. Methods Nine patients diagnosed with oral cancer were enrolled. One radiation oncologist, 1 head and neck surgeon, and 9 oral surgeons performed image evaluations individually. Interobserver variations in tumor size were estimated by using magnetic resonance imaging (MRI) scans as controls. The 11 observers scored each CT/MRI scan in terms of clarity and quality. IDCT and NRCT scans were scored in terms of artifact of the image, internal structure, and boundary with other organs, and total number of images for diagnosis. The 11 observers scored on a scale of 1 to 5 for the above-listed items. A subjective assessment of image quality was performed by scoring on a scale of 1 to 5, with 1 indicating no detection of the desired anatomy or features, 3 indicating adequate image quality, and 5 indicating higher-than-needed image quality. Results The mean score for artifact of the image was significantly higher with IDCT than with NRCT among 3 of the 11 observers. The mean score for internal structure was significantly higher with DECT than with NRCT among 5 of the 11 observers. The mean score for boundary with other organs was significantly higher with IDCT than with NRCT among 2 of the 11 observers. Finally, the mean score for total number of images for diagnosis was significantly higher with IDCT than with NRCT among 5 of the 11 observers. The comparison of tumor size showed no significant difference among NRCT, IDCT, and MRI. However, the size evaluation by MRI was larger than that by CT, and the interobserver variation was small. Conclusions Visual estimation is superior with IDCT than with NRCT in terms of all the aspects. IDCT can be recommended for evaluating oral cancer.