Investigate The Factors Influencing Lung Nodules Detected By CT

Purpose :Investigate the factors influencing lung nodules being detected, so we can control these factors to reduce the un-detected lung nodules.Material and method: We divide our research into three parts. The first part: Investigate the factors influencing lung nodules being detected by standard exposing dose. Select 21 patients with lung metastasis to scan by CT with different schemes, including different slice thickness (6. 5mm and 3. 2mm) , different reconstruction interval (0% and 50% overlapping-reconstructed) and different pitch(1.25 and 0.875). After scaning, measure the size, number and position of those nodules detected by different schemes , respectivily, then evaluate the correlativity of scan schemes , nodule size, nodule position and the number of nodule detected. The second part investigate the influence of different exposing dose to detecting lung nodules. 26 patients with lung metastasis were selected to scan by CT with defferent protocols, including 195 mAs /6.5mm (exposing dose/scan thickness), 195 /3. 2, 50 /3. 2, 30 /3. 2 in all. Then add up the number , size , position of nodules detected with dfferent dose respectivly. The third part: Investigate the influence of different exposing dose to the quality of image. 26 patients with lung metastasis were selected to scan by CT with differentschemes, including 195 mAs /6. 5mm (exposing dose/scan thickness), 195 /3. 2, 50 /3. 2, 30 /3. 2. Then evaluate the correlation among exposing dose, slice thickness, body height/weight, the clearity of lung-structure displayed and noise of image.Result:The first is part: The lung nodules detected by 3.2mm slice thickness and 50%-overlapping reconstructed are significantly more than 6.5mm and 0% respectively; the main superiority of former is the higher ability detecting nodules smaller than 5mm and 2mm than later. Andthe nodules closing together pneumo-vessels, hilar of lung, para - mediastina and diaphram are more undetected than the nodules not closing together those structure. But the number of nodules detected by pitch 0.875 and pitch 1.25 has no significanct difference. The second part: The numbers of lung nodules detected with 195 /3.2, 50 /3. 2 and 30 /3.2 have no significant difference, but when classificated by their size, as the exposing dose decreased, nodules smaller than 2mm or 5mm are detected decreasingly, morever, the noise of image increases. Although the image-noise of 30/3.2 is significant higher than 195/6. 5,the number of nodules detected by former is significant more than later. The three part: The noise of 30 /3. 2 is bigger than 50 /3. 2, and 50 /3. 2 bigger than 195 /3. 2. As the exposing dose decreased, the clearity of image decreases. The quality of 50 /3. 2 may satisfy the require of viewing and diagnosis, even is more superio than 195/6.5 in displaying some structure. Howerer, the quality of 30 /3. 2 decreases so greatly that it influences viewing for some objects with body heigh/Weight bigger than 2.513.Conclusion: Exposing dose, slice thickness, reconstruction interval, size of the nodule, density of the nodule, location of the nodule all may influence the lung nodule being detected. Increasing the slice thickness, adding the reconstruction interval, and the nodule being too small or closing together pneumo-vessels, hilum of lung, para-mediastina and diaphram all can make some nodules being un-detected; in additional, decreasing exposing dose, the artifact of image may obscure small nodules with low density. In all these factors,slice thickness and size of the nodule are the two most important facters. Decreasing the slice thickness can make up for the loss of scanning with low exposing dose, and scanning by low exposing dose with thin slice thickness is superior to scanning by standard dose with routine slice thickness. Auto-selecting exposing dose by the height/weigth is benefit to control the quality.