The purpose of this paper is to develop a new algorithm to decrease the number of false-positive candidates which were obtained with a previous detection scheme, based on a genetic-algorithm (GA), for lung nodules in digital chest radiography. The algorithm consists of the following three steps. Firstly, the candidates with the marginal area just outside the lung were eliminated by the histogram analysis. Secondly, the Gaussian operator was used to enhance the bone edge of the candidates. After analyzing the kurtosis of orientation-amplitude histogram and entropy of interpolative difference signal of central region, the candidates including the shadows of bones were eliminated. Thirdly, the modified Laplacian-Gaussian operator which suppresses the high-frequency noise and the low-frequency components was used to enhance the nodules. After this filtering, the candidate area was divided into three rectangular regions (R1, R2 and R3). Three features, which consist of the entropy of interpolative difference signal in R1, the ratio of variances between R2 and R1, and that between R3 and R 1, were used to eliminate falsepositive candidates of the shadows of vessels and others. The 341 candidates (7 true-positive and 334false-positive candidates) from 20 clinical chest images (10 normal and 10 abnormal cases) were analyzed and 297 false-positive ones were eliminated by our method. In other words, we were able to decrease the number of false-positive candidates from approximately 17 to 1.9 per image without reducing any true nodules.

抄録全体を表示

We are developing an automated detection scheme for the clustered microcalcifications in digital mammograms. This scheme is part of an overall package for computer-aided diagnosis (CAD) system. When different cases (databases) are used for testing, it is required to readjust many parameters of the scheme for these new cases. One of the solutions to facilitate this procedure is proposed mainly for the case that the imaging characteristic in terms of contrast is different. By measuring the contrast of microcalcifications from selected samples, the relationship between the contrast and the background pixel value was determined, and then the contrast correction curve was produced. The curve for the second case was set to adjust the contrast difference between the two databases to match the first case. We demonstrated that our technique by use of contrast correction curve was effective when 32 mammograms from a different institution were tested.

抄録全体を表示

In this paper, the discrimination of ultrasonic heart (echocardiographic) images is studied by making use of texture features which are obtained from a gray-level cooccurrence matrix. Features of these types are used as inputs to the input layer of a neutral network (NN) to discriminate three sets of echocardiographic imagesnormal heart, dilated cardiomyopathy (DCM), and hypertrophic cardiomyopathy (HCM) (18,13, Vol.13 No.2 (1996) 61 and 6 samples, respectively). The performance of the NN classifier is compared to that of a minimum distance (MD) classifier and fractal dimension (FD). Our results show that the NN produces about 83%, the MD and FD produce about 76% and 70% correct discrimination (three outputs), respectively. These results indicate that the method of feature-based image analysis using the NN has potential utility for computer-aided diagnosis of the DCM, HCM, and other heart diseases.

抄録全体を表示

Computer-generated random number is a kind of pseudo-random number. So far various tests of random numbers have been performed. However, comparisons of random numbers generated by different programming languages have not been made. Generally, random numbers are often used when simulation is conducted using a computer. Therefore it is important to compare the variation of random numbers resulting from different programming languages used. In the present study, the comparison among three types of programming languages, namely, the C, Pascal and the Basic, is made by calculating number π using Monte Carlo method.

抄録全体を表示

We are developing a computer-aided diagnostic system (CAD) which can detect masses and clustered microcalcifications on digital mammograms. In indicating the computer results to the physicians on an ordinary cathode-ray tube monitor, an automated windowing method has been applied to enhance the shadow of mammary glands. The area near the breast skin lines was difficult to evaluate, because most of pixel values in those areas were ignored by the method. Therefore, the improvement of contrast-correction process is required for the CAD. The purpose of this paper is to describe the dynamic-range compression technique which could make observers easier to recognize the regions in high and/or low density area, and is also to apply the method to the display stage of annotated images in our mammogram CAD.

抄録全体を表示