Chooi W K, Matthews S, Bull M J, Morcos S K
Department of Diagnostic Imaging, Northern General Hospital, Sheffield S5 7AU, UK.
Br J Radiol. 2003 Aug;76(908):536-40. doi: 10.1259/bjr/21813112.
We examined 23 consecutive patients (11 males and 12 females with mean age of 56 years) with possible airway diseases to assess the impact of multiplanar image reconstruction (MPR) on the degree of confidence and accuracy in diagnosing bronchial abnormalities and emphysema. The thorax was scanned contiguously at 1 mm slice thickness using Siemens Volume Zoom Multislice CT scanner. Images were reconstructed at 1 mm slice thickness (lung windows L-600HU W-1600HU utilizing high spatial frequency algorithm) in the axial (10 mm apart), sagittal (4 images per lung) and coronal (6 images) plane. Paddle wheel image reconstructions were also performed in the assessment of bronchiectasis. Axial images were assessed with and without the help of MPR by three chest radiologists at two separate occasions (at least 4 weeks apart). The presence of bronchiectasis, emphysema and bronchiolitis in each lobe was documented on a confidence scale of 0 to 3. The overall mean confidence for each observer with and without MPR was compared. Consensus diagnosis was used as the gold standard for the assessment of the diagnostic accuracy of each observer. A confidence score of 2 or more for any lobe was considered diagnostic of the particular airway disease. The diagnostic accuracy for each observer with and without MPR was compared. Consensus reporting diagnosed bronchiectasis in 7 patients (30.4%), bronchiolitis in 5 patients (21.7%) and emphysema in 12 patients (52%). MPR did not increase the confidence of assessing the different abnormalities for all observers but improvement in diagnosing bronchiectasis was noted in two observers. The improvement did not reach statistical significance. However, agreement between observers in the diagnosis of bronchiectasis and emphysema was improved when the MPR images were used in conjunction with standard axial imaging (Kappa statistic improved from 0.29 to 0.54 for bronchiectasis and from 0.7 to 0.81 for emphysema). Agreement on the diagnosis of bronchiolitis was not improved by MPR for all observers. Our results suggest that MPR seems to improve the confidence in diagnosing bronchiectasis and emphysema.
我们检查了23例连续的可能患有气道疾病的患者（11名男性和12名女性，平均年龄56岁），以评估多平面图像重建（MPR）对诊断支气管异常和肺气肿的置信度及准确性的影响。使用西门子容积变焦多层CT扫描仪以1毫米层厚连续扫描胸部。图像以1毫米层厚重建（肺窗L - 600HU W - 1600HU，采用高空间频率算法），包括轴向（间隔10毫米）、矢状面（每侧肺4幅图像）和冠状面（6幅图像）。在评估支气管扩张时还进行了桨轮图像重建。三位胸部放射科医生在两个不同时间（至少间隔4周）分别在有无MPR帮助的情况下评估轴向图像。每个肺叶支气管扩张、肺气肿和细支气管炎的存在情况以0至3的置信度等级记录。比较了每位观察者在有和无MPR情况下的总体平均置信度。共识诊断用作评估每位观察者诊断准确性的金标准。任何肺叶的置信度评分达到2或更高被认为可诊断为特定的气道疾病。比较了每位观察者在有和无MPR情况下的诊断准确性。共识报告诊断出7例支气管扩张患者（30.4%）、5例细支气管炎患者（21.7%）和12例肺气肿患者（52%）。MPR并没有提高所有观察者评估不同异常的置信度，但两名观察者在诊断支气管扩张方面有改善。这种改善未达到统计学意义。然而，当MPR图像与标准轴向成像结合使用时，观察者之间在支气管扩张和肺气肿诊断上的一致性得到了改善（支气管扩张的Kappa统计量从0.29提高到0.54，肺气肿从0.7提高到0.81）。对于所有观察者，MPR并未改善细支气管炎诊断的一致性。我们的结果表明，MPR似乎提高了诊断支气管扩张和肺气肿的置信度。
