2010 年 66 巻 8 号 p. 870-878
In optimizing exposures, it is very important to evaluate the impact of image noise on image quality. To realize this, there is a need to evaluate how much image noise will make the subject disease invisible. But generally it is very difficult to shoot images of different quality in a clinical examination. Thus, a method to create a noise addition image by adding the image noise to raw data has been reported. However, this approach requires a special system, so it is difficult to implement in many facilities. We have invented a method to easily create a noise addition image by using the water phantom and image add-subtract software that accompanies the device. To create a noise addition image, first we made a noise image by subtracting the water phantom with different SD. A noise addition image was then created by adding the noise image to the original image. By using this method, a simulation image with intergraded SD can be created from the original. Moreover, the noise frequency component of the created noise addition image is as same as the real image. Thus, the relationship of image quality to SD in the clinical image can be evaluated. Although this method is an easy method of LDSI creation on image data, a noise addition image can be easily created by using image addition and subtraction software and water phantom, and this can be implemented in many facilities.
