Analysis of Breast Thermograms using Coherence Enhanced Diffusion Filter and Radon Transform

摘要

Breast thermography plays a significant role in early detection of breast cancer by quantifying thermal variations associated with precancerous state of breast. The distribution of asymmetrical thermal patterns in breast images demonstrates the presence of abnormalities. In this work, breast thermal images are analysed using coherence enhanced diffusion filter based denoising and radon transform based structural texture features. The breast tissues are extracted from background tissue by multiplying ground truth masks with denoised images. The midpoint of inframammary folds is identified to separate left and right regions from segmented images. Normal and abnormal groups are categorized based on the healthy and pathological conditions of the separated breast tissues. Radon transform based structural texture features such as coarseness, contrast directionality are extracted from denoised and raw images. This filter could perform simultaneous noise reduction and edge preservation at different orientations. The signal to noise ratio of denoised images is found to be enhanced by 40% when compared to that of raw images. The effect of denoising enhanced the difference in feature values of normal and abnormal breast tissues by 17%. Radon transformation could effectively enhance the structural information in an image. Features such as contrast and directionality extracted from denoised images show distinct variation between normal and abnormal subjects and are effective in detecting pathological conditions present in breast tissues. Coherence enhanced diffusion filter technique could effectively enhance the signal to noise ratio thereby preserving clear edges. Radon transform based structural texture features are found to be distinct in differentiating normal and carcinoma tissues. Hence, it appears that the proposed method could help in enhancing the diagnostic relevance of breast thermograms.