Improved image quality of low-dose thoracic CT examinations with a new postprocessing software.

In 2008 a phantom study indicated that there is a potential for reducing the CT doses when using a new postprocessing filter. The purpose of this study was to test this new postprocessing filter clinically for low-dose chest CT examinations, to assess whether the diagnostic performance is the same or improved. A standardized clinical chest CT protocol was used on patients with colorectal cancer. Only mA settings changed between patients according to patient size. One standard and one low-dose chest protocol were performed for all patients. The low-dose images were postprocessed with a new software filter, which provides context-controlled restoration of digital images by using adaptive filters. Three radiologists assessed randomly all the images independently. A total of 24 scan series were evaluated with respect to image quality according to quality criteria from the European guidelines for chest CT using a five-point scale; 576 details were assessed. Overall mean score is the average score for all details rated for all three readers for all full-dose series, low-dose series and low-dose enhanced series, respectively. The statistical methods used for comparison were paired sampled t-test and intraclass correlation coefficient. The postprocessing filter improved the diagnostic performance compared to the unenhanced low-dose images. Mean score for full-dose, low-dose and low-dose enhanced series were 3.8, 3.0 and 3.3, respectively. For all patients the full-dose series gave higher scores than the low-dose series. Intraclass correlation coefficients were 0.2, 0.1 and 0.3 for the full-dose, low-dose and low-dose enhanced series, respectively. There is a potential for improving diagnostic performance of low-dose CT chest examinations using this new postprocessing filter.