Biomedical Engineering Branch, Division of Basic & Applied Sciences, National Cancer Center, 111 Jungbalsan-ro, Ilsandong-gu, Goyang-si, Gyeonggi-do, Goyang, South Korea.
J Digit Imaging. 2011 Jun;24(3):424-36. doi: 10.1007/s10278-010-9273-x.
Whole-body bone scan is one of the most frequent diagnostic procedures in nuclear medicine. Especially, it plays a significant role in important procedures such as the diagnosis of osseous metastasis and evaluation of osseous tumor response to chemotherapy and radiation therapy. It can also be used to monitor the possibility of any recurrence of the tumor. However, it is a very time-consuming effort for radiologists to quantify subtle interval changes between successive whole-body bone scans because of many variations such as intensity, geometry, and morphology. In this paper, we present the most effective method of image enhancement based on histograms, which may assist radiologists in interpreting successive whole-body bone scans effectively. Forty-eight successive whole-body bone scans from 10 patients were obtained and evaluated using six methods of image enhancement based on histograms: histogram equalization, brightness-preserving bi-histogram equalization, contrast-limited adaptive histogram equalization, end-in search, histogram matching, and exact histogram matching (EHM). Comparison of the results of the different methods was made using three similarity measures peak signal-to-noise ratio, histogram intersection, and structural similarity. Image enhancement of successive bone scans using EHM showed the best results out of the six methods measured for all similarity measures. EHM is the best method of image enhancement based on histograms for diagnosing successive whole-body bone scans. The method for successive whole-body bone scans has the potential to greatly assist radiologists quantify interval changes more accurately and quickly by compensating for the variable nature of intensity information. Consequently, it can improve radiologists' diagnostic accuracy as well as reduce reading time for detecting interval changes.
全身骨扫描是核医学中最常见的诊断程序之一。特别是，它在重要程序中发挥着重要作用，如骨转移的诊断和评估骨肿瘤对化疗和放疗的反应。它还可用于监测肿瘤复发的可能性。然而，由于强度、几何形状和形态等多种变化，放射科医生对连续全身骨扫描之间细微的间隔变化进行定量是一项非常耗时的工作。在本文中，我们提出了基于直方图的最有效的图像增强方法，这可能有助于放射科医生有效地解释连续的全身骨扫描。从 10 名患者中获得了 48 次连续的全身骨扫描，并使用基于直方图的 6 种图像增强方法进行了评估：直方图均衡化、亮度保持双直方图均衡化、对比度限制自适应直方图均衡化、末端搜索、直方图匹配和精确直方图匹配（EHM）。使用峰信噪比、直方图交叠和结构相似性三种相似性度量对不同方法的结果进行了比较。在所有相似性度量中，EHM 对连续骨扫描的图像增强效果最好。EHM 是诊断连续全身骨扫描的最佳基于直方图的图像增强方法。该方法用于连续全身骨扫描，通过补偿强度信息的可变性，有可能极大地帮助放射科医生更准确、快速地量化间隔变化。因此，它可以提高放射科医生的诊断准确性，减少检测间隔变化的阅读时间。
