Yu-Che Cheng

Robust ellipse detection based on hierarchical image pyramid and Hough transform.

In this research we propose a fast and robust ellipse detection algorithm based on a multipass Hough transform and an image pyramid data structure. The algorithm starts with an exhaustive search on a low-resolution image in the image pyramid using elliptical Hough transform. Then the image resolution is iteratively increased while the candidate ellipses with higher resolution are updated at each step until the original image resolution is reached. After removing the detected ellipses, the Hough transform is repeatedly applied in multiple passes to search for remaining ellipses, and terminates when no more ellipses are found. This approach significantly reduces the false positive error of ellipse detection as compared with the conventional randomized Hough transform method. The analysis shows that the computing complexity of this algorithm is Θ(n(5/2)), and thus the computation time and memory requirement are significantly reduced. The developed algorithm was tested with images containing various numbers of ellipses. The effects of noise-to-signal ratio combined with various ellipse sizes on the detection accuracy were analyzed and discussed. Experimental results revealed that the algorithm is robust to noise. The average detection accuracies were all above 90% for images with less than seven ellipses, and slightly decreased to about 80% for images with more ellipses. The average false positive error was less than 2%.

Partial concordance between nuclear and organelle DNA in revealing the genetic divergence among Quercus glauca (Fagaceae) populations in Taiwan

Quercus glauca (Thunb. ex Murray) Oerst (Fagaceae) has a wide distributional range in Taiwan. In this study, the evolutionary history and the most genetically divergent sites of Q. glauca were studied using a nuclear gene marker, glyceraldehyde‐3‐phosphate dehydrogenase. Also, the consistency of the results obtained from nuclear gene and cytoplasmic loci was investigated. Using a genealogical approach (TCS software), we determined haplotypes and their relationships to one another. We used the level of divergence for each population from the remaining populations (calculated as mean values of pairwise population differentiation, FST, for each population) to locate the most genetically divergent areas in Taiwan. According to the average FST of each population in comparison with the remaining ones, a peak was found in the northern part of central Taiwan, and another was found in the southeastern region. The peak profiles of the mean FST values for all three DNA data sets (nDNA, cpDNA, and mtDNA) showed similar trends on both sides of the Central Mountain Ridge, except for the mtDNA sequence on the western side. This study suggests that two potential refugia existed in Taiwan during the last glaciation: one in the northern part of central Taiwan and another in southern Taiwan.

A post-processing method for correction and enhancement of chemical shift images

Chemical shift imaging (CSI) relies on a strong and homogeneous main field. Field homogeneity ensures adequate coherence between the precessions of individual spins within each voxel. Variation of field strength between different voxels causes geometric distortion and intensity variation in chemical shift images, resulting in errors when analyzing the spatial distribution of specific chemical compounds. A post-processing method, based on detection of the spectral peak of water and baseline subtraction with Lorentzian functions, was developed in this study to automatically correct spectra offsets caused by field inhomogeneity, thus enhancing the contrast of the chemical shift images. Because this method does not require prior field plot information, it offers advantages over existing correction methods. Furthermore, the method significantly reduces geometric distortion and enhances signals of chemical compounds even when the water suppression protocol was applied during the CSI data acquisition. The experimental results of the water and glucose phantoms showed a considerable reduction of artifacts in the spectroscopic images when this post-processing method was employed. The significance of this method was also demonstrated by an analysis of the spatial distributions of sugar and water contents in ripe and unripe bananas.

Chemical Shift Imaging of Spatial-Temporal Changes of Lycopene in Tomatoes During Ripening

Ripening is an important process for fruits to be more attractive and to create more economic value in the market. For tomatoes, their colors turn from green to red during ripening. This obvious external color change is associated with the accumulation of lycopene. Lycopene is a long-chain unsaturated cartenoid that is beneficial to human health. To gain a better understanding of the synthesis of lycopene during ripening, MRI (Magnetic resonance imaging) technique was employed in this study. CSI (chemical shift imaging) is a branch of MRI technique which provides both spatial and spectroscopic information of the test sample. The spatial structure can be shown as MR images while the chemical properties can be revealed with NMR (nuclear magnetic resonance) spectroscopy. By calculating the intensity of the spectrum of each voxel, the spatial distribution of a specific chemical compound can be displayed in a chemical shift image. This study aims to develop an analytical method to detect the spatial changes of lycopene content in tomatoes during ripening process by using CSI technique. An algorithm was first developed to correct the image artifact due to field inhomogeneity, and thus to visualize the spatial distribution of lycopene internal of the same tomato before and after ripening. CSI experiments were performed to acquire tomato images with the same imaging parameters at the mature green stage and the red ripe stage. After comparing the relative distribution area of lycopene content in different ripening stages, the analysis showed that the distribution area of lycopene content increased significantly when the tomato’s during ripening. When comparing different parts of a tomato, the increase of lycopene content was most manifest in the outer pericarp part. To support the findings from the CSI analyses, HPLC (high performance liquid chromatography) experiments were also carried out to compare the lycopene contents of tomatoes at the mature green stage and the red ripe stage. The results of the HPLC experiments were consistent with that of the CSI analyses.

Physico-Chemical Analysis of Internal Bruise of Selected Fruits Using Chemical Shift Imaging

Chemical shift imaging (CSI) is a spectroscopic imaging technique which provides both spatial and spectroscopic information of internal physico-chemical properties of test samples and, thus, it is a useful tool for assessing the distribution of internal characteristics of fruits or vegetables. In this study, CSI was applied in detecting the internal properties of bruises in apples and plums subjected to mechanical impact. MR spectroscopic correction and image fusion were employed in correcting field inhomogeneity and displaying CSI experimental results. The result images reflected the variation of some phenolic compounds and water content in the bruised and fleshy parts of the fruits on the first day and the seventh day after bruising. The quantity of phenolic compounds in the bruised part in apples and plums decreased after 7 days, but it in the fleshy part increased after 7 days. Water content in apple has the similar result of phenolic compounds. It reduced in bruised part after 7 days later. HPLC experiments were applied to support the CSI results. In HPLC analyses, phenolic compounds would increase in bruised and fleshy parts in apples and plums by the 7th day, and on the 7th day, phenolic compounds in fleshy part in apples were more than that in the bruised part. However, phenolic compounds in the bruised part of plums were more than that in the fleshy part. The results of 7th day between CSI and HPLC are consistent. This study demonstrates that CSI is a feasible method for non-invasive assessment of internal physico-chemical properties of biological materials.

Analyses of Ripening Process in Tomatoes Using Chemical Shift Magnetic Resonance Imaging

Magnetic resonance imaging (MRI) has been widely applied in biological research and diagnostics. The chemical shift magnetic resonance imaging (CSI) technique is a spectroscopy imaging technique in which spatially localized spectra are obtained simultaneously from a volume of the sample. This technique allows for the determination of chemical composition as well as the spatial information of the sample and therefore it is an efficient tool to assess internal characteristics of fruits and vegetables. This paper presents the experimental results and analyses of chemical shift imaging of tomatoes of different ripening stages. The major physico-chemical components such as water content, sugar content, and lycopene which are related to the ripening process of tomatoes were examined for their spatial-temporal distribution in individual tomatoes. To analyze the transient changes of each physico-chemical components, a spectral signal processing procedure and software program were developed to visualize the dynamics of tomato ripening process with pseudo-color mapping. The efficacy of CSI is demonstrated with quantitative analyses of the internal sugar content distribution of tomato during ripening process.

In Vivo Measurement of Human Tympanic Membrane Structure

Abstract Problem: To noninvasively quantify the in vivo human tympanic membrane (TM) structure for further explaining some of the variability in long-term hearing improvement after tympanoplasty surgery for chronic otitis media. Methods: All MR images of 5 female and 5 male volunteers’ TM, semicircular canals, and cochlea were acquired by a 3-T MR system (Brucker S630) using a surface coil of 9 cm diameter and 3D SNAP sequences, aged 20 to 35 years. The pixel resolution of the acquired image for the volunteers’ right ears was 0.59 × 0.63 × 0.47 mm by 3D SNAP. A 3D solid model of in vivo TM, semicircular canals, and cochlea was then reconstructed based on the acquired MR images. The geometric characteristics of TM, semicircular canals, and cochlea were subsequently determined by 2 indices including surface area and volume. Results: The surface areas were 108.34, 152.94, and 108.15 square mm, and the volumes were 36.47, 46.02, and 30.18 cubic mm for TM structure of 3 volunteers, respectively. The preliminary data indicated that small surface area and large volume would obtain the thick TM. Variability of dimensional characteristics of 7 more volunteers’ TM was further analyzed. The number of volunteers did not reach a statistically significant level. Conclusion: The dimensional characteristics of in vivo human TM was quantified by using surface area and volume indices. The thickness of TM was then qualified by surface area and volume. Using the proposed measurement method, the spatial configuration of the in vivo TM can be tracked pre- and postoperation noninvasively. Significance: Knowledge of the structure of TM may help determine the cause of continued problems such as the age-related changes in the mechanics of the TM, and can explain part of the age-related hearing loss observed. Some patients with surgically repaired tympanic membrane perforation still have hearing loss. Support: None reported.

Chemical Shift Magnetic Resonance Imaging for Spectroscopic Analysis of Biological Tissues

Magnetic resonance imaging (MRI) has been widely applied in biological research and diagnostics. The chemical shift magnetic resonance imaging technique is a sequence of matrix with one spectroscopic and two spatial dimensions. This technique allows for the determination of chemical composition as well as the spatial information of the specimen and therefore it is an efficient tool to assess internal characteristics of various biological tissues. This paper introduces the process of acquiring chemical shift images of selected biological tissues and the determinations of imaging parameters for spectral images of optimum quality. An image analysis software was developed and image analysis algorithms such as image segmentation, pseudo-color processing, and image fusion were established for the analysis of spectral images. The software allows for displaying the acquired data either as a set spectrum representing the chemical composition of each pixel, or image slices representing the spatial distribution of an individual component. The efficacy of chemical shift imaging is demonstrated in this research with quantitative analysis of lean and fat portions of animal tissues.