Abstract:
Quantitative contrast-enhanced breast computed tomography (CT) has the potential to improve the diagnosis and management of breast cancer. Traditional CT methods using energy-integrated detectors and dual-exposure images with different incident spectra for material discrimination can increase patient radiation dose and be susceptible to motion artifacts and spectral resolution loss. Photon Counting Detectors (PCDs) offer a promising alternative approach, enabling acquisition of multiple energy levels in a single exposure and potentially better energy resolution. Gallium arsenide (GaAs) is particularly promising for breast PCD-CT due to its high quantum efficiency and reduction of fluorescence x-rays escaping the pixel within the breast imaging energy range. In this study, the spectral performance of a GaAs PCD for quantitative iodine contrast-enhanced breast CT was evaluated. A GaAs detector with a pixel size of 100μm, a thickness of 500μm was simulated. Simulations were performed using cylindrical phantoms of varying diameters (10 cm, 12 cm, and 16 cm) with different concentrations and locations of iodine inserts, using incident spectra of 50, 55, and 60 kVp with 2 mm of added aluminum filtration and and a mean glandular dose of 10 mGy. We accounted for the effects of beam hardening and energy detector response using TIGRE CT open-source software and the publicly available Photon Counting Toolkit (PcTK). Material-specific images of the breast phantom were produced using both projection and image-based material decomposition methods, and iodine component images were used to estimate iodine intake. Accuracy and precision of the proposed methods for estimating iodine concentration in breast CT images were assessed for different material decomposition methods, incident spectra, and breast phantom thicknesses. The results showed that both the beam hardening effect and imperfection in the detector response had a significant impact on performance in terms of Root Mean Squared Error (RMSE), precision, and accuracy of estimating iodine intake in the breast. Furthermore, the study demonstrated the effectiveness of both material decomposition methods in making accurate and precise iodine concentration predictions using a GaAs-based photon counting breast CT system, with better performance when applying the projection-based material decomposition approach. The study highlights the potential of GaAs-based photon counting breast CT systems as viable alternatives to traditional imaging methods in terms of material decomposition and iodine concentration estimation, and proposes phantoms and figures of merit to assess their performance.
摘要:
定量对比增强乳房计算机断层扫描(CT)具有改善乳腺癌诊断和治疗的潜力。使用能量集成检测器和具有不同入射光谱的双曝光图像进行材料辨别的传统方法会增加患者辐射剂量,并且容易受到运动伪影和光谱分辨率损失的影响。光子计数探测器(PCD)提供了一种有前途的替代方法,能够在一次曝光中获取多个能量水平,并可能获得更好的能量分辨率。砷化镓(GaAs)由于其高量子效率和减少了在乳房成像能量范围内逸出像素的荧光X射线而特别有希望用于乳房PCD-CT。在这项研究中,评估了GaAsPCD用于定量碘对比增强乳腺CT的光谱性能。

像素尺寸为100μm的GaAs检测器,模拟了500μm的厚度。使用不同直径的圆柱形体模进行模拟(10厘米,12厘米,和16厘米),具有不同的碘插入物浓度和位置,使用50、55和60kVp的入射光谱,添加2mm的铝过滤和对应于约10mGy的平均腺体剂量(MGD)的一个暴露水平。我们使用TIGRECT开源软件和公开可用的光子计数工具包(PcTK)考虑了光束硬化和能量检测器响应的影响。使用投影和基于图像的材料分解方法产生乳房的特定材料图像,和碘成分图像用于估计碘摄入量。针对不同的材料分解方法,评估了预测乳腺CT图像中碘浓度的方法的准确性和准确性。入射光谱,和胸模厚度。

结果表明,光束硬化效应和检测器响应中的缺陷都对均方根误差(RMSE)方面的性能产生了重大影响,精度,和估计的准确性。