背景ノイズを抑制した肺結節明瞭化法

抄録

We have ever proposed pulmonary nodule clarity method in chest radiographs by controlling the pulmonary vessel (or its en face view) which is frequently extracted as false-positives. However, insufficient suppression of the clavicle, rib and peripheral pulmonary vessel shadows made difficult to detect true lung nodules. In this study, density alteration of the clavicle, rib and peripheral pulmonary vessel shadows was defined as back ground noise. We propose a new method to control back ground noise and further clarify pulmonary nodules by applying wavelet analysis and error diffusion method to our previous proposed pulmonary nodule clarity method. A radiologist evaluated the new images obtained by the proposed technique in a point of suppression of the back ground noise and visibility of the pulmonary nodules. In this evaluation of 117 images excluding "extremely subtle" and "obvious" from 154 images with nodules in the JSRT database, the back ground noise was sufficiently suppressed in 80.2％. The visibility of the pulmonary nodules was well improved in 88.0％, comparing with the previous pulmonary nodule clarity images. It became easy to find true lung nodules by suppression of the back ground noise. The proposed method is expected to be useful for detection of lung cancer nodules.