Mir A H, Hanmandlu M, Tandon S N
Centre for Biomedical Engineering, IIT Delhi, Hauz Khas, India.
Front Med Biol Eng. 1996;7(2):93-109.
Diagnostic features in computed tomography (CT) images vary widely due to the vague nature (fuzziness) of functional characteristics in organ pathologies. Classical image enhancement techniques cannot adapt to the characteristics of this nature. A method based on fuzzy logic is given in this paper. In this method the image of interest is transformed into a fuzzy plane using fuzzifiers which can be changed to select a crossover point. A contrast intensification operator (INT) is then applied. The operator increases the grade of membership of those values which lie above the crossover point and decreases the same for those pixels which lie below it, thereby increasing the contrast and reducing the fuzziness. Quantitative measures of fuzziness which reflect a sort of quantitative measure of image quality have also been studied. Using digitized CT images it has been shown that at early stages of a disease, when the contrast of the pathological tissues is very low, the visibility of the disease could be considerably improved using this technique.
由于器官病变中功能特征的模糊性,计算机断层扫描(CT)图像中的诊断特征差异很大。传统的图像增强技术无法适应这种模糊性质的特征。本文给出了一种基于模糊逻辑的方法。在该方法中,使用模糊化器将感兴趣的图像转换为模糊平面,模糊化器可以改变以选择一个交叉点。然后应用对比度增强算子(INT)。该算子增加位于交叉点上方的那些值的隶属度,并降低位于交叉点下方的那些像素的隶属度,从而增加对比度并减少模糊性。还研究了反映图像质量某种定量度量的模糊性定量度量。使用数字化CT图像已经表明,在疾病的早期阶段,当病理组织的对比度非常低时,使用该技术可以显著提高疾病的可见性。