Carl E. Ravin Advanced Imaging Laboratories, Department of Radiology, Duke University Medical Center, Durham, North Carolina 27705, USA.
Med Phys. 2013 Sep;40(9):091916. doi: 10.1118/1.4818060.
Stereoscopic chest biplane correlation imaging (stereo∕BCI) has been proposed as an alternative modality to single view chest x-ray (CXR). The metrics effective modulation transfer function (eMTF), effective normalized noise power spectrum (eNNPS), and effective detective quantum efficiency (eDQE) have been proposed as clinically relevant metrics for assessing clinical system performance taking into consideration the magnification and scatter effects. This study compared the metrics eMTF, eNNPS, eDQE, and detectability index for stereo∕BCI and single view CXR under isodose conditions at two magnifications for two anthropomorphic phantoms of differing sizes.
Measurements for the eMTF were taken for two phantom sizes with an opaque edge test device using established techniques. The eNNPS was measured at two isodose conditions for two phantoms using established techniques. The scatter was measured for two phantoms using an established beam stop method. All measurements were also taken at two different magnifications with two phantoms. A geometrical phantom was used for comparison with prior results for CXR although the results for an anatomy free phantom are not expected to vary for BCI.
Stereo∕BCI resulted in improved metrics compared to single view CXR. Results indicated that magnification can potentially improve the detection performance primarily due to the air gap which reduced scatter by ∼20%. For both phantoms, at isodose, eDQE(0) for stereo∕BCI was ∼100 times higher than that for CXR. Magnification at isodose improved eDQE(0) by ∼10 times for stereo∕BCI. Increasing the dose did not improve eDQE. The detectability index for stereo∕BCI was ∼100 times better than single view CXR for all conditions. The detectability index was also not improved with increased dose.
The findings indicate that stereo∕BCI with magnification may improve detectability of subtle lung nodules compared to single view CXR. Results were improved with magnification for the smaller phantom but not for the larger phantom. The effective DQE and the detectability index did not improve with increasing dose.
立体胸部双平面相关成像（立体/BCI）已被提议作为替代单视图胸部 X 射线（CXR）的方法。有效调制传递函数（eMTF）、有效归一化噪声功率谱（eNNPS）和有效量子探测效率（eDQE）已被提议作为评估临床系统性能的临床相关指标，同时考虑放大率和散射效应。本研究比较了两种不同尺寸的体模在两种放大率下的立体/BCI 和单视图 CXR 在等剂量条件下的 eMTF、eNNPS、eDQE 和检测指数。
使用不透明边缘测试装置对两个体模的 eMTF 进行了测量，采用了已建立的技术。使用已建立的技术，在两个体模的两个等剂量条件下测量了 eNNPS。使用已建立的光束截止方法测量了两个体模的散射。所有测量均在两个不同的放大率下进行，使用了两个体模。使用几何体模进行比较，尽管对于 BCI 来说，无解剖结构体模的结果不应有所不同，但对于 CXR 来说，其结果是可以预期的。
与单视图 CXR 相比，立体/BCI 的指标得到了改善。结果表明，放大率可以潜在地提高检测性能，主要是由于空气间隙减少了约 20%的散射。对于两个体模，在等剂量下，立体/BCI 的 eDQE(0)比 CXR 高约 100 倍。立体/BCI 的放大率在等剂量下提高了 eDQE(0)约 10 倍。增加剂量并不能提高 eDQE。在所有条件下，立体/BCI 的检测指数比单视图 CXR 好约 100 倍。增加剂量也不能提高检测指数。
研究结果表明，与单视图 CXR 相比，具有放大率的立体/BCI 可能提高对细微肺结节的检测能力。对于较小的体模，放大率会提高结果，但对于较大的体模则不会。有效 DQE 和检测指数不会随着剂量的增加而提高。
