Noise characteristics of the reduction image displayed on liquid crystal display in digital mammography

Soft-copy diagnosis of medical images is currently widespread. Because the pixel size of a digital mammogram is very small, the matrix size is extremely large. Especially the matrix size of phase contrast mammography (PCM) is very large compared with a conventional mammography. When such an image is displayed on a liquid crystal display (LCD), it is displayed as a reduction image. Therefore, it is necessary to use an appropriate reduction rate and an interpolation method such that the reduction processing does not influence the diagnosis. We obtained a uniform image exposure and measured the noise power spectrum (NPS) of the image reduced by using the nearest neighbor, bilinear, and bicubic methods with several reduction rates. Our results showed that the best interpolation method was the bilinear method. Moreover, the NPS value increased by a factor of the square of the inverse of the reduction rate.