Tai-Been Chen

Segmentation of cDNA microarray images by kernel density estimation

The segmentation of cDNA microarray spots is essential in analyzing the intensities of microarray images for biological and medical investigation. In this work, nonparametric methods using kernel density estimation are applied to segment two-channel cDNA microarray images. This approach groups pixels into both a foreground and a background. The segmentation performance of this model is tested and evaluated with reference to 16 microarray data. In particular, spike genes with various contents are spotted in a microarray to examine and evaluate the accuracy of the segmentation results. Duplicated design is implemented to evaluate the accuracy of the model. The results of this study demonstrate that this method can cluster pixels and estimate statistics regarding spots with high accuracy.

EMPIRICAL STUDIES OF CROSS-REFERENCE MAXIMUM LIKELIHOOD ESTIMATE RECONSTRUCTION FOR POSITRON EMISSION TOMOGRAPHY

Positron emission tomography (PET) can reveal subtle metabolic process, which is an important modality for diagnosis. However, spatial resolution of PET images is not as good as computed tomography (CT) or magnetic resonance imaging (MRI), which can show precise anatomical details. Our study is to improve image quality of PET using better reconstruction methods. In this paper, we use a new and efficient method to incorporate the correlated structural information obtained from MRI. A mean estimate smoothing the maximum likelihood estimate (MLE) locally within each region of interest is derived according to the boundaries provided by the structural information. Since the boundaries may not be correct, a penalized MLE using the mean estimate is sought. The resulting reconstruction is called a cross-reference maximum likelihood estimate (CRMLE). The CRMLE is obtained through a modified expectation maximization (EM) algorithm, which is shown to be computationally efficient by our phantom and clinical studies.

String-averaging expectation-maximization for maximum likelihood estimation in emission tomography

We study the maximum likelihood model in emission tomography and propose a new family of algorithms for its solution, called string-averaging expectationmaximization (SAEM). In the string-averaging algorithmic regime, the index set of all underlying equations is split into subsets, called ‘strings’, and the algorithm separately proceeds along each string, possibly in parallel. Then, the end-points of all strings are averaged to form the next iterate. SAEM algorithms with several strings present better practical merits than the classical row-action maximum-likelihood algorithm. We present numerical experiments showing the effectiveness of the algorithmic scheme, using data of image reconstruction problems. Performance is evaluated from the computational

Quantifying Liver Cirrhosis by Extracting Significant Features from MRI T2 Image

Most patients with liver cirrhosis must undergo a series of clinical examinations, including ultrasound imaging, liver biopsy, and blood tests. However, the quantification of liver cirrhosis by extracting significant features from a T2-weighted magnetic resonance image (MRI) provides useful diagnostic information in clinical tests. Sixty-two subjects were randomly selected to participate in this retrospective analysis with assigned to experimental and control groups. The T2-weighted MRI was obtained and to them dynamic adjusted gray levels. The extracted features of the image were standard deviation (SD), mean, and entropy of pixel intensity in the region of interest (ROI). The receiver operator characteristic (ROC) curve, 95% confidence intervals, and kappa statistics were used to test the significance and agreement. The analysis of area under ROC shows that SD, mean, and entropy in the ROI were significant between the experimental group and the control group. Smaller values of SD, mean, and entropy were associated with a higher probability of liver cirrhosis. The agreements between the extracted features and diagnostic results were shown significantly (P < 0.001). In this investigation, quantitative features of SD, mean, and entropy in the ROI were successfully computed by the dynamic gray level scaling of T2-weighted MRI with high accuracy.

Investigated geometrical characteristics and image density of left ventricle of multi-detector computed tomography in early coronary artery disease patients

BACKGROUND Coronary artery disease (CAD) remains the leading cause of death worldwide. Currently, cardiac multi-detector computed tomography (MDCT) is widely used to diagnose CAD. The purpose in this study is to identify informative and useful predictors from left ventricular (LV) in the early CAD patients using cardiac MDCT images. MATERIALS AND METHODS Study groups comprised 42 subjects who underwent a screening health examination, including laboratory testing and cardiac angiography by 64-slice MDCT angiography. Two geometrical characteristics and one image density were defined as shape, size and stiffness on MDCT image. The t-test, logistic regression, and receiver operating characteristic curve were applied to assess and identify the significant predictors. The Kappa statistics was used to exam the agreements with physicians judgments (i.e., Golden of True, GOT). RESULTS The proposed three characteristics of LV MDCT images are important predictors and risk factors for the early CAD patients. These predictors present over 80% of AUC and higher odds ratio. The Kappa statistics was 0.68 for the combinations of shape and stiffness into logistic regression. CONCLUSIONS The shape, size and stiffness of the left ventricular on MDCT can be used to be the effective indicators in the early CAD patients. Besides, the combinations of shape and stiffness into logistic regression could provide substantial agreement with physicians judgments.

64-bit and 128-bit DX random number generators

Extending 32-bit DX generators introduced by Deng and Xu (ACM Trans Model Comput Simul 13:299–309, 2003), we perform an extensive computer search for classes of 64-bit and 128-bit DX generators of large orders. The period lengths of these high resolution DX generators are ranging from 101915 to 1058221. The software implementation of these generators can be developed for 64-bit or 128-bit hardware. The great empirical performances of DX generators have been confirmed by an extensive battery of tests in the TestU01 package. These high resolution DX generators can be useful to perform large scale simulations in scientific investigations for various computer systems.

Diagnosed chest lesion on diffusion-weighted magnetic resonance images using apparent diffusion coefficients.

PURPOSE A novel diagnostic method using the standard deviation (SD) value of apparent diffusion coefficient (ADC) by diffusion-weighted (DWI) magnetic resonance imaging (MRI) is applied for differential diagnosis of primary chest cancers, metastatic tumors and benign tumors. MATERIALS AND METHODS This retrospective study enrolled 27 patients (20 males, 7 female; age, 15-85; mean age, 68) who had thoracic mass lesions in the last three years and underwent an MRI chest examination at our institution. In total, 29 mass lesions were analyzed using SD of ADC and DWI. Lesions were divided into five groups: Primary lung cancers (N = 10); esophageal cancers (N = 5); metastatic tumors (N = 8); benign tumors (N = 3); and inflammatory lesions (N = 3). Quantitative assessment of MRI parameters of mass lesions was performed. The ADC value was acquired based on the average of the entire tumor area. The error-plot, t-test and the area under receiver operating characteristic (AUC) were applied for statistical analysis. RESULTS The SD of ADC value (mean±SD) was (4.867±1.359)×10-4 mm2/sec in primary lung cancers, and (3.598±0.350)×10-4 mm2/sec in metastatic tumors. The SD of ADC values of primary lung cancers and metastatic tumors (P <  0.05) were significantly different and the AUC was 0.800 (P <  0.05). The means of SD of ADC values was 4.532±1.406×10-4 mm2/sec and 2.973±0.364×10-4 mm2/sec for malignant tumors (including primary lung cancers, esophageal cancers) and benign tumors with respectively. The mean of SD of ADC values between malignant chest tumors and benign chest tumors was shown significant difference (P <  0.01). The values of AUC was 0.967 between malignant chest tumors and benign chest tumors (P <  0.05). The ADC values for primary lung cancers, metastatic tumors and benign tumors were not significantly difference (P >  0.05). CONCLUSIONS The mean of SD of ADC value by DWI can be used for differential diagnosis of chest lesions.

Effects of pelvic floor muscle exercise on faecal incontinence in rectal cancer patients after stoma closure

The purpose of this study was to examine the effects of pelvic floor muscle exercise (PFME) on the faecal incontinence (FI) of rectal cancer patients following stoma closure. Participants were randomly distributed into an exercise group (n = 27) and non-exercise group (n = 26). An experimental design and longitudinal approach were implemented for data collection. Baseline data were collected at 1 day before discharge, and then PFME was taught before the patients were discharged from the hospital. We collected data and followed up with the patients at their pre-discharge visit and at 1, 2, 3, 6 and 9 months after discharge. The Cleveland Clinic Faecal Incontinence (CCI) score was used to measure patient outcome. PFME proved to effectively decrease the degree of FI in stoma closure recipients. The FI score of the exercise group significantly decreased from 8.37 to 2.27 after PFME compared with that of the non-exercise group (from 8.54 to 2.58). The generalised estimation equation tests showed that both group and time were significantly different. The tests also indicated that although PFME appeared to hasten the decline of incontinence, this effect was no longer detectable at 9 months; thus, it may be an effective intervention for FI when implemented up to half a year after discharge.

Analyzed outcome for embolization of hepatic carcinoma by C-arm angiographic computed tomography.

PURPOSE This study evaluated and monitored the outcome of angiographic embolization of hepatic carcinoma by real-time C-arm angiographic computed tomography under number of tumors, size of tumors, and patients age.METHODS AND MARTIALS: In total, 142 patients underwent angiographic embolization of hepatic carcinoma. The control group, 71 patients, underwent conventional angiographic (CA) embolization of hepatic carcinoma. The experimental group, 71 patients, underwent C-arm angiographic computed tomography (CCT) embolization of hepatic carcinoma. The numbers of angiographic embolization, number of tumors, size of tumors, and patients ages were recorded for comparisons between groups by analysis of variance (ANOVA) with cross-interaction and the chi-square test (cross table). RESULTS The age ranges were 20-84 and 35-84 years old for the experimental and control groups respectively. Average number of angiographic embolizations of hepatic carcinomas were 2.63 ± 1.84 and 5.32 ± 2.01 for the experimental and control groups. The number of angiographic embolizations under number of tumors, size of tumors, and patients ages between groups were significantly different (P< 0.05). The effective analyses of transcatheter arterial chemoembolization (TACE) by CCT were significant by chi-square test (P< 0.05) under ⩽ 3 cm and patients aged ⩽ 60. CONCLUSION The main advantage by CCT for undergoing TACE under tumor size smaller than 3 cm and numbers of tumor smaller 3 times were more significantly effective than those by CA. The CCT combined with TACE had high potentially reduced numbers of undergoing TACE.

Iterative image reconstruction with random correction for PET studies

A maximum likelihood-expectation maximization (ML-EM) reconstruction algorithm has been developed that allows random coincidence correction for the phantom we used and the reconstructed images are better than those obtained by convolution backprojection (CBP) for positron emission tomography (PET) studies in terms of spatial resolution, image artifacts and noise. With our algorithm reconstruct the true coincidence events and random coincidence events were reconstructed separately. We also calculated the random ratio from the measured projection data (singles) using line and cylindrical phantoms, respectively. From cylindrical phantom experiments, the random event ratio was 41.8% to 49.1% in each ring. These results are close to the ratios obtained from geometric calculation, which range from 45.0% to 49.5%. The random ratios and the patterns of random events provide insightful information for random correction. This information is particularly valuable when the delay window correction is not available as in the case of our PET system.