シグモイド関数型ウェーブレット展開係数変換曲線による医用画像のコントラスト強調

抄録

Pre-processing is important for digital medical images because it can contribute to the improvement of diagnostic accuracy with medical images and computer-aided diagnosis. In this paper, a novel sigmoid-type transfer curve for wavelet coefficient weighting adjustment was proposed to improve contrast of medical images. The proposed method was applied to chest radiographs, mammograms and chest CT images, and then its performance was quantitatively evaluated with contrast improvement ratio. As a result, contrasts of all applied images were obviously improved. Compared with the conventional exponential-type transfer curve, the proposed sigmoid-type transfer curve was more effective with a statistical significance(p<0.05).