[in Japanese]

View full abstract

In this article, I attempted to apply an evolutionary concept based on the selfish gene and the media as the extension of man to medical information systems. After introducing some definitions such as gene, meme, media, communication and information, I proposed the principle of media-success. Then, I showed that the computerized patient record (CPR)could become popular as a new powerful media of health insurers.

View full abstract

Reinforcement learning (RL) is an approach to machine intelligence. It combines the fields of dynamic programming and supervised learning to yield powerful machine-learning systems. The RL appeals to many researchers because of its generality. However, it has not been used yet in the field of image processing. In RL, the computer is simply given a goal to achieve. The computer then learns how to achieve that goal by trial-and-error interactions with its environment. Of the RL methods Q-learning is a typical learning approach. In this paper, we present a novel method for image segmentation based on the Q-learning. Additionally, we illustrate the proposed algorithm and demonstrate its effectiveness for image contour extraction and region-of-interest detection using three medical images. Our preliminary results are promising.

View full abstract

We have proposed a new method for detection of organ contours in medical images. Our method is modified active contour model by replacement of edge information of image with region information, such as average, variance of pixel values and stochastic property based on Markov random field, which allow us to introduce rules of image property. And we have introduced the technique to split the contour, which has an intersection due to noise or small organs in images. We applied this method to three-dimensional (3-D) computed tomography (CT) images and compared with results by previous methods. As a result, proposed method detected the organ contour more correctly than the previous methods.

View full abstract

We have attempted to calculate a primary X-ray spectrum of CT (Computed Tomography)system. Dose rates of primary X-ray spectrum are very high, so we reduced dose rate by using carbon scatterer and enabled to measure 90°-scattered spectrum with a CdZnTe detector. The primary X-ray spectrum can be calcuIated from measured data of 90°-scattered spectrum by counting backwards. In counting backwards, we used not only Klein-Nishina coefficients but also response functions obtained by Monte Carlo methods, because Raylei scattering and multiple scattering maybe occur in carbon scatterer. In our results, the primary X-ray spectrum calculated from response functions obtained by Monte Carlo methods are roughly equivalent to that calculated from Klein-Nishina coefficients.

View full abstract