[Detection method for blurred regions in radiographs].

In this paper, we propose a detection method for blurred regions in radiographs. The method involves edge detection using a Sobel filter, manually determining the region of interest (ROI), feature calculation, and classification using a support vector machine. We applied our method to 14 phantom images (7 normal images, 7 blurred images) and 14 clinical images (12 normal images, 2 blurred images). As a result, the average classification accuracies of ROIs with blurring and ROIs without blurring were 98% and 90% for phantom images and clinical images, respectively.

A modified undecimated discrete wavelet transform based approach to mammographic image denoising.

In this work, the authors present an effective denoising method to attempt reducing the noise in mammographic images. The method is based on using hierarchical correlation of the coefficients of discrete stationary wavelet transforms. The features of the proposed technique include iterative use of undecimated multi-directional wavelet transforms at adjacent scales. To validate the proposed method, computer simulations were conducted, followed by its applications to clinical mammograms. Mutual information originating from information theory was used as an evaluation measure for selection of an optimal wavelet basis function. We examined the performance of the proposed method by comparing it with the conventional undecimated discrete wavelet transform (UDWT) method in terms of processing time-consuming and image quality. Our results showed that with the use of the proposed method the computation time can be reduced to approximately 1/10 of the conventional UDWT method consumed. The results of visual assessment indicated that the images processed with the proposed UDWT method showed statistically significant superior image quality over those processed with the conventional UDWT method. Our research results demonstrate the superiority and effectiveness of the proposed approach.

Computerized detection of lung nodules by CT for radiologic technologists in preliminary screening.

In Japan, radiologists and radiologic technologists are endeavoring to improve the quality of lung CT screening. In particular, preliminary screening by radiologic technologists is expected to decrease radiologists' burden and improve the accuracy of CT screening. We considered that an application of computer-aided detection (CAD) would also be as useful in preliminary screening as in the radiologist's regular reading. Our purpose in this study was to investigate the potential of the application of CAD to preliminary screening. CAD software that we developed was applied to 17 lung CT scans that radiologic technologists had pre-interpreted. A radiologist recognized 29 lung nodules from the CT images, whereas radiologic technologists did not recognize 11 of the 29 nodules at their pre-reading. Our CAD software detected lung nodules at an accuracy of 100% (29/29), with 4.1 false positives per case. The 11 nodules that radiologic technologists did not recognize were included in the CAD-detected nodules. This result suggests that the application of CAD may aid radiologic technologists in their preliminary screening.

[Ultrasonographic study of the hip joint in infants].

We performed ultrasonography screening for the hip joint in 145 infants between January 2003 and January 2004 at Nakajo-machi in Niigata Prefecture. This ultrasonographic study was conducted by Graf's method. The method could discriminate normal hip joints and was able to visualize 11 abnormal hip joints as well. All abnormal hip joints were dysplastic. This method is more useful than x-ray examination for infants, since the cartilaginous femoral head and the soft tissues around the hip could be clearly visualized. As a result, ultrasonography appears to be a useful screening method for congenital dislocation of the hip joint, without the hazard of radiation caused by x-rays.