Diagnostic value of MIP imaging: Detection of small pulmonary nodules with MIP imaging in Computed Tomography

Since the pulmonary nodules can be the precursors of lung cancer in lung patologies, early and accurate diagnosis is of great importance.Nowadays, direct radiography remains prior imaging modalitiy in diagnosis of pulmonary nodules. After direct radiography, computed tomographic (CT) evaluation is the most commonly used and most sensitive imaging modality in diagnosis and follow up of pulmonary nodules. Apart from this basic method Positron Emission Tomography (PET) and Magnetic Resonance İmaging (MRI) for detecting pulmonary nodules are used less frequently. In radiology, particularly the difficulty of the detection of small pulmonary nodules constitute one of the majör problems. Today, pulmonary nodules followed by CT may be overlooked during the review depending on their sizes,intensity or localization. With CT there may ocur failures due to technical reasons during the detection of especially below 1 cm nodules. In order to diagnose small pulmonary nodules with high sensitivity and speed,various reserches have been made on CT techniques. Among them two independent reading, slice thickness reduction, three dimentional reconstruction technique such as maximum intensity projection (MIP) imaging and volume rendering (VR) and a new CT based nodüle diagnosis technique which is called computer aided design (CAD) can be located. There are many studies and arguments about the role of MIP technique on the detection of lung nodules. The aim of this study, by examining the patients which have small pulmonary nodules, is to identifiy the improvement in sensitivity and speed of MIP technique at nodule detection compared to classical methods. For this purpose following the approval of the ethic committee between April 2012 and December 2012 patients who admited to Başkent University Hospital with several complaints and in the radiology departmant the lung nodules detected with thorax CT examination of 69 patients (339 nodules) were included in study. 3 mm thick axial sections of patients with MIP section in axial plane taken at different times was evaluated by two separately radiologists keeping time without considering previous reports. In statistical analysis, the correlation between the MIP images and the classical 3 mm axial images which are called as "gold standart" was evaluated using "intraclass correlation coefficient". In addition, whether the reporting time difference between the groups is meaningful or not was determined by using "Paired Sample T-Test". MIP technique sensitivity for each radiologist in detection of nodules below 5 mm and between 5-10 mm are calculated separately.Harmony among radiologists was evaluated using "intraclass correlation coefficient". As a result of the statistical analysis harmony evaluation between the radiologists was determined as 86.8%, p<0.05 which is statsticaly significant.With intraclass corelation coefficient test, the complience of the MIP evaluation of the first radiologist with the gold standart are 78.6% for smaller than 5 mm nodules and 84.7% for between 5-10 mm nodules.For the second radiologist, these values become 90.3% and 84.4 % respectively. All of these values are p<0.05 and harmony between the groups was statistically significant. Sensitivities of the first and second radiologists in detecting nodules with the MIP technique were found 81.4% and 83.4% respectively. It can be started, by comparing the evaluation time period of MIP images with the gold standart, that reporting time periods for both radiologist decreased significantly. As a result the use of MIP images, as an alternative method for detection of lung nodules, shortened the assesment time significantly and the nodule numbers detected in MIP images are close to the gold standard axial images and the MIP images show high sensitivity. MIP technique may preferred due to the fact that according to these data the evaluation of lung nodules with MIP technique is fast and reliable to provide information.