ABSTRACT
Objective::To construct the color matching template irrelevant to size and rotation according to the types and distribution characteristics of colors in images of Chinese herbal pieces, in order to establish color characterization and image identification methods for Chinese herbal pieces. Method::Totally 20 types of Chinese herbal pieces were selected, including rhizomes, flowers, seeds and fruits.For each sample, two observation surfaces were selected to extract color parameters in foreground through image processing such as image segmentation, model transformation from RGB to L*a*b*.Color vectors of the two observation surfaces were sequenced in a descending order, scaled to a certain size by interpolating, and combined into an integrated color vector in a weight ratio of 1∶1.As for centripetally distributed observation surface images(e.g.transverse section), corrosion operation was conducted to extract the color components of each ring from outer to inner by circles, which were then ordered and scaled.The integrated color vector was used as initial template for training, the correlation coefficient between each sample and the template was calculated, and the interval estimation of positive samples were carried out by t test.With the total recognition rate as an indicator, the optimal template dimensions, width of ring and training volume were ultimately determined. Result::The visualization results of the trained templates of the varied herbal pieces were easy to be visually distinguished.After 260 samples of the herbal pieces were tested, the template of a and b components was better than that of L*, a* and b* in terms of recognition performance, with a* recognition accuracy of 95.8%(249/200). Conclusion::Color characteristics of images from two observation surfaces of Chinese herbal pieces are integrated to obtain the combined color feature vector, so as to achieve preferable recognition results for samples from both the same and different medicinal parts.This method boasts a strong anti-interference ability of random variation of sample shape, sampling part and color.
ABSTRACT
Objective:To build a gray-level matching template by using the gray level information of the microscopic image of the transverse section of Chinese medicinal materials,in order to realize the automatic recognition of the images of Chinese medicinal materials independent of scale and orientation. Method:By using the embedding method of polyethylene glycol (PEG),the transverse slices of 19 kinds of common rhizomatous medicinal materials were obtained. The images of the slices were taken by digital microscopic imaging technology,and the mosaic grayscale images were obtained by image registration,noise removal and boundary location. The center of the structure of the materials in the images was selected to establish the polar coordinate system, so as to divide grids from the radial and angular directions. By counting the gray information in each grid,the gray information digital matrix that can characterize the microscopic identification characteristics of the materials was obtained. Images in an appropriate sample size was used to train the matrix for generalization of the matrix. The covariance coefficients between the matrix of positive or negative verification sample and the template matrix were calculated to set the best identification parameters. For each medicinal material,80 fan-shaped images were prepared,including 70% of training samples,15% of validation samples and 15% of test samples,and single template and template set were tested with test samples. Result:In the test of 240 images including non-template-set medicinal materials,the correct recognition rate of single-template test was 90.1%,and that of template-set test was 92.5%. Conclusion:This method can well characterize the microscopic identification characteristics of Chinese medicinal materials, with a strong anti-interference ability and less subjective-errors, acquire sample images easily, and provide technical support for the digitization of morphological quality control of Chinese medicinal materials.
ABSTRACT
The ultrasound Doppler fetal heart rate measurement is the gold standard of fetal heart rate counting. However, the existing fetal heart rate extraction algorithms are not designed specifically to suppress the high maternal interference during the second stage of labor, and false detection occurrences are common during labor. With this background, a method combining time-frequency frame template library optimal selecting and non-linear template matching is proposed. The method contributes a template library, and the optimal template can be selected to match the signal frame. After the short-time Fourier transform of the signal, the difference between the signal and the template is optimized by leaky rectified linear unit (LReLU) function frame by frame. The heart rate was calculated from the peak of the matching curve and the heart rate was calculated. By comparing the proposed method with the autocorrelation method, the results show that the detection accuracy of the proposed method is improved by 20% on average, and the non-linear template matching of 23% samples is at least 50% higher than the autocorrelation method. This paper designs the algorithm by analyzing the characteristics of the interference and signal mixing. We hope that this paper will provide a new idea for fetal heart rate extraction which not only focuses on the original signal.
Subject(s)
Female , Humans , Pregnancy , Algorithms , Fetal Monitoring , Fourier Analysis , Heart Rate, Fetal , Signal Processing, Computer-Assisted , Ultrasonography, DopplerABSTRACT
Objective: To introduce a new template peak matching algorithm for calculating the similarity of chromatographic fingerprint of traditional Chinese herbs. Methods: Tanshinone II A in S. miltiorrhiza Bge of different batches were determined by HPLC and the chromatograms of them were obtained. We designated a new peak matching algorithm based on the previous algorithms, which employed a certain real chromatographic fingerprint as their template. In the new algorithm, we arranged the retention times of chromatographic peaks of all chromatograms into an ascending order, forming a template. The sorting procedure complied with the following 2 rules. First, the same peak matches the nearest corresponding peak. Second, one peak does not appear twice in the same chromatogram. Results: The new algorithm avoided the shortcomings of previous algorithm, such as mismatching and missed matching, and obtained a satisfactory outcome. Conclusion: Our new algorithm provides a basis for improving the reliability of retention time-based peak matching algorithm.
ABSTRACT
OBJECTIVE: The purpose of this study was to develop a new method for automated lung nodule detection in serial section CT images with using the characteristics of the 3D appearance of the nodules that distinguish themselves from the vessels. MATERIALS AND METHODS: Lung nodules were detected in four steps. First, to reduce the number of region of interests (ROIs) and the computation time, the lung regions of the CTs were segmented using Genetic Cellular Neural Networks (G-CNN). Then, for each lung region, ROIs were specified with using the 8 directional search; +1 or -1 values were assigned to each voxel. The 3D ROI image was obtained by combining all the 2-Dimensional (2D) ROI images. A 3D template was created to find the nodule-like structures on the 3D ROI image. Convolution of the 3D ROI image with the proposed template strengthens the shapes that are similar to those of the template and it weakens the other ones. Finally, fuzzy rule based thresholding was applied and the ROI's were found. To test the system's efficiency, we used 16 cases with a total of 425 slices, which were taken from the Lung Image Database Consortium (LIDC) dataset. RESULTS: The computer aided diagnosis (CAD) system achieved 100% sensitivity with 13.375 FPs per case when the nodule thickness was greater than or equal to 5.625 mm. CONCLUSION: Our results indicate that the detection performance of our algorithm is satisfactory, and this may well improve the performance of computer-aided detection of lung nodules.
Subject(s)
Humans , Algorithms , Automation , Diagnosis, Computer-Assisted , False Positive Reactions , Fuzzy Logic , Imaging, Three-Dimensional , Lung Neoplasms/diagnostic imaging , Neural Networks, Computer , Radiographic Image Interpretation, Computer-Assisted , Sensitivity and Specificity , Tomography, X-Ray ComputedABSTRACT
Objective To explore a method for real-time pretreatment and character extraction of epicardial signals and promote the research on the electrophysiological mechanism of atrial fibrillation (AF).Methods Firstly,an analysis was made to get the properties of epicardial signals. Then,a method based on template matching was used to extract atrial activities from epicardial signals. At last,the feature points of atrial activities could be recognized.Results The correctness of recognizing feature points of sinus rhythm and AF were 99.7% and 97.0% respectively. Furthermore,only 1~2 s were cost to compute a 10-second data record with this method.Conclusion The proposed method can extract crucial information about atrium depolarization from epicardial signals and is adequate for online analysis.