This paper proposes an automated procedure for segmenting menisci in MR images of a human knee aided by fuzzy expert system. A three-dimensional (3D) MR volumetric images composed of many slice images consists of several parts: bone marrow, meniscus, periarticular liquor, cartilage and others. We employ both T1-weighted and T2-weighted MR images to identify the menisci with high accuracy. After a registration between these images is manually done on a computer display, our procedure aided by fuzzy expert system can automatically segment meniscal regions from 3D MR images. Physicians can observe the 3D shapes of meniscal tears from any point of view on the display. We examined five subjects including a normal knee and three injured knees. The all meniscal regions were significantly identified, and these 3D shapes were displayed. The patterns of meniscal tears were identified on the display for all subjects. In a subject, since the preoperative and postoperative 3D meniscal shapes were clearly viewed, we easily recognized the operated meniscal regions. Thus, the system can provide useful information for diagnosing meniscal tears.
Conventional depth estimation methods for light field cameras often fail to estimate sharp object boundaries. We propose a depth estimation method using a wavelet-based matching cost calculation and an estimation optimization, which can estimate sharp object boundaries. Our method first calculates matching costs on four subband images derived with a wavelet transform, and then combines the costs based on matching confidence. After an initial depth estimation by using the matching cost, we improve estimation by an optimization process. Experimental results demonstrated that our proposed method provides more plausible estimation with sharper object boundaries.
In recognition using SIFT and SURF, matching of features extracted from both input images and learning images are involved. However, the recognition target is not always facing the same direction as that of the learning image. For this reason, recognition is performed by making it learn the image of the object you wish to recognize by capturing it from a wide variety of angles. Also, where repeated patterns are included in the object to be recognized, mismatching and misrecognition often occur because of the fact that the repeated pattern feature descriptors are similar. To avoid this issue, similarity of feature descriptors is measured using cosine similarity, and features with a specific degree of similarity are excluded. In particular, as this degree of similarity differs depending on the image, in this paper, we propose a method of seeking the optimal degree of similarity depending on the image and excluding repeated pattern features. Finally, we verify the effectiveness of the proposed method by showing some experimental results.
This paper presents a novel method for reconstructing the temperature distributions of water around a small heated sphere and predicting the heat generation rates of the sphere. This method is based on the temperature dependence of the absorbance of water at the wavelength of 1150 nm. Absorbance images at 1150 nm were obtained when a 1 mm diameter steel sphere located in water contained in a glass cell with a light path length of 10 mm was heated by a 760 kHz alternating magnetic field. Inverse Abel transform is applied to the line profiles of the absorbance, and radial temperature distributions are reconstructed. The heat generation rates calculated from these temperature distributions well agree with induction heating power levels.
A prediction method of internal temperatures distributed axisymmetrically in an aqueous solution with a thickness from sub-mm to a few mm is presented. This method consists of a near-infrared (NIR) absorption imaging and an inverse Abel transform. In the NIR absorption imaging, absorbance images at the wavelength of 1412 nm, the most temperature-sensitive wavelength in the v_1 + v_3 absorption band of water, are obtained by using a narrow-bandpass filter and an NIR camera. In the case that the internal temperature possesses an axisymmetric distribution, it can be predicted by Abel inversion of the measured absorbance profile. In this study the absorbance profiles are approximated by multi-Gaussian functions because their inverse transform can be calculated analytically. Cross-sectional temperature distributions around a thin hot wire in 1.5 mm thick water are shown.
This paper describes a design method of automated medical diagnosis system (AMDS), which provides a normal degree for a disease. In this paper, we consider blood test data of human. Suppose a disease which can be diagnosed by the test, first, we do statistical analysis. Second, we determine the reference range of the test. Third, we design a fuzzy inference system consisting of membership functions based on the reference range. The inference system plays primary role in the AMDS. Finally, we show the design of AMPS for diabetes and the experimental results.
