Environmental Science & Technology Theses and Dissertations
Permanent URI for this collectionhttp://hdl.handle.net/1903/2748
Browse
11 results
Search Results
Item Advanced imaging and data mining technologies for medical and food safety applications(2009) Jiang, Lu; Tao, Yang; Biological Resources Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)As one of the most fast-developing research areas, biological imaging and image analysis receive more and more attentions, and have been already widely applied in many scientific fields including medical diagnosis and food safety inspection. To further investigate such a very interesting area, this research is mainly focused on advanced imaging and pattern recognition technologies in both medical and food safety applications, which include 1) noise reduction of ultra-low-dose multi-slice helical CT imaging for early lung cancer screening, and 2) automated discrimination between walnut shell and meat under hyperspectral florescence imaging. In the medical imaging and diagnosis area, because X-ray computed tomography (CT) has been applied to screen large populations for early lung cancer detection during the last decade, more and more attentions have been paid to studying low-dose, even ultra-low-dose X-ray CTs. However, reducing CT radiation exposure inevitably increases the noise level in the sinogram, thereby degrading the quality of reconstructed CT images. Thus, how to reduce the noise levels in the low-dose CT images becomes a meaningful topic. In this research, a nonparametric smoothing method with block based thin plate smoothing splines and the roughness penalty was introduced to restore the ultra-low-dose helical CT raw data, which was acquired under 120 kVp / 10 mAs protocol. The objective thorax image quality evaluation was first conducted to assess the image quality and noise level of proposed method. A web-based subjective evaluation system was also built for the total of 23 radiologists to compare proposed approach with traditional sinogram restoration method. Both objective and subjective evaluation studies showed the effectiveness of proposed thin-plate based nonparametric regression method in sinogram restoration of multi-slice helical ultra-low-dose CT. In food quality inspection area, automated discrimination between walnut shell and meat has become an imperative task in the walnut postharvest processing industry in the U.S. This research developed two hyperspectral fluorescence imaging based approaches, which were capable of differentiating walnut small shell fragments from meat. Firstly, a principal component analysis (PCA) and Gaussian mixture model (PCA-GMM)-based Bayesian classification method was introduced. PCA was used to extract features, and then the optimal number of components in PCA was selected by a cross-validation technique. The PCA-GMM-based Bayesian classifier was further applied to differentiate the walnut shell and meat according to the class-conditional probability and the prior estimated by the Gaussian mixture model. The experimental results showed the effectiveness of this PCA-GMM approach, and an overall 98.2% recognition rate was achieved. Secondly, Gaussian-kernel based Support Vector Machine (SVM) was presented for the walnut shell and meat discrimination in the hyperspectral florescence imagery. SVM was applied to seek an optimal low to high dimensional mapping such that the nonlinear separable input data in the original input data space became separable on the mapped high dimensional space, and hence fulfilled the classification between walnut shell and meat. An overall recognition rate of 98.7% was achieved by this method. Although the hyperspectral fluorescence imaging is capable of differentiating between walnut shell and meat, one persistent problem is how to deal with huge amount of data acquired by the hyperspectral imaging system, and hence improve the efficiency of application system. To solve this problem, an Independent Component Analysis with k-Nearest Neighbor Classifier (ICA-kNN) approach was presented in this research to reduce the data redundancy while not sacrifice the classification performance too much. An overall 90.6% detection rate was achieved given 10 optimal wavelengths, which constituted only 13% of the total acquired hyperspectral image data. In order to further evaluate the proposed method, the classification results of the ICA-kNN approach were also compared to the kNN classifier method alone. The experimental results showed that the ICA-kNN method with fewer wavelengths had the same performance as the kNN classifier alone using information from all 79 wavelengths. This demonstrated the effectiveness of the proposed ICA-kNN method for the hyperspectral band selection in the walnut shell and meat classification.Item RESPIRATORY MECHANICS OF FLOW LIMITATION AND CHARACTERIZATION OF RESISTANCE MEASUREMENTS WITH A NON-INVASIVE DEVICE(2009) Coursey, Derya; Johnson, Arthur T; Biological Resources Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Resistance measurements with the airflow perturbation device (APD) were compared to directly measured pulmonary resistances with an esophageal balloon to validate the APD. The APD perturbs the flow and the mouth pressure during regular breathing. The ratio of mouth pressure perturbations to the flow perturbations was used to calculate the inspiratory, expiratory and average respiratory resistance. Six healthy subjects were tested during tidal breathing when known external resistances were added during inspiration, during expiration, and during both inspiration and expiration. The difference between the averaged APD measured and directly measured pulmonary resistances was 0.59 ± 1.25 (mean ± SD) cmH2O/L/s. Compared to the magnitude of the known increase in added resistance, the APD measured resistance increased by 79 %, while the directly measured pulmonary resistance increased only by 56%. During addition of external resistances to both inspiration and expiration, the changes in inspiratory and expiratory pulmonary resistance were only 36 % and 62 % of the added resistance, respectively. On the other hand, the APD inhalation and exhalation resistance measured between 82 % and 76 % of added resistance change. It was concluded that the APD detects changes in external resistance at least as well and probably better than classical measurements of pulmonary resistance. Additionally, expiratory isovolume pressure - flow (IVPF) curves, which show the pressure at which the flow becomes limited during forced expiration, were constructed in six healthy subjects with the classical invasive method of esophageal balloon (EB) and the alternative noninvasive method of stop - flow (SF) at 25, 50, and 75 % vital capacity (VC). The difference between the pressures (Pmax) and flow (Qmax) at which flow limitation first occurs and correlation with the stop - flow and esophageal balloon methods were studied. Additionally, the resistance at flow limitation was compared to the APD resistance during forced breathing. On average, PSF,max was 5.6 and 4.4 times PEB,max at 25 %VC and 50 %VC, respectively. QSF,max was 0.68 and 0.59 times QEB,max at 25 %VC and 50 %VC, respectively. No correlation was found between the stop - flow and esophageal balloon methods as well as between the resistances at flow limitation.Item PROSPECTIVE HEAD MOVEMENT CORRECTION FOR HIGH-RESOLUTION MRI USING AN IN-BORE OPTICAL TRACKING SYSTEM(2009) Qin, Lei; Tao, Yang; Biological Resources Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)In MRI of the human brain, subject motion is a major cause of magnetic resonance image quality degradation. To compensate the effects of head motion during data acquisition, an in-bore optical motion tracking system is proposed. The system comprises one or two MR compatible infrared cameras that are fixed on a holder right above and in front of the head coil. The resulting close proximity of the cameras to the object allows precise tracking of its movement. During image acquisition, the MRI scanner uses this tracking information to prospectively compensate for head motion by adjusting gradient field direction and RF phase and frequency. Experiments performed on subjects demonstrate the system's robustness, exhibiting an accuracy of better than 0.1mm and 0.15˚.Item Noninvasive Optical Imaging Techniques as a Quantitative Analysis of Kaposi's Sarcoma Skin Lesions(2007-11-26) Vogel, Abby Jeanne; Tao, Yang; Biological Resources Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)The visible inspection and palpation of skin lesions have long been used to assess the course of cutaneous disease in individuals with Kaposi's sarcoma (KS). Assessing the KS lesions requires a highly trained evaluator and evaluations made by different observers or by the same observer at different times can be inconsistent. Since lesions can grow from slow to explosively fast, and be associated with mortality and morbidity, reliable assessment of the lesions is important. Optical imaging techniques are quantitative methods that potentially offer a more objective means of assessing skin health that can supplement visual clinical observations. In this dissertation, the first paper describes using thermography and laser Doppler imaging (LDI) to monitor the temperature and vasculature, respectively, of KS lesions undergoing experimental therapy. Results showed that the median temperature elevation of lesions compared to the surrounding tissue was 1.1 ºC (range -0.68 to 3.43 ºC). In addition, 12 of the 16 lesions studied had increased blood perfusion as assessed by LDI (median 66 arbitrary units (AU), range -44 to 451 AU). The second paper describes the use of near-infrared (NIR) multi-spectral imaging to provide functional information about the lesions and surrounding tissue. Multi-spectral images were input to a mathematical optical skin model based on the absorption and scattering properties of skin, including the effects of melanin, blood, and oxygenated and deoxygenated hemoglobin. Functional information about the lesions before and after treatment with experimental therapies was determined. The third paper describes Monte-Carlo simulations of tissue conducted to determine the detection limits of a typical multi-spectral imaging system. The results showed that biological information contained in a typical spectral image reflected a small volume of tissue situated vertically under each pixel from a depth less than 2-3 mm. The objects appearing on a spectral image reasonably reproduce the correct geometrical shape and size of underlying inclusions of pathological tissue. Based on the three papers included in this dissertation, these three imaging techniques were found to be objective, easy to perform, and appear to be very sensitive in quantitatively assessing KS lesion progress upon administration of therapy.Item The Effect Of Inspiratory Air Humidity And Temperature On Performance Time While Wearing A Respirator(2006-05-09) Francis, Erica; Johnson, Arthur T; Biological Resources Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Respirators are worn in only about 20-30% of the appropriate circumstances. This research examined the effect of inspired air conditions on performance time. An environmental chamber supplied air at 27°C, 28°C, 37°C, and 55°C to a powered air-purifying respirator (PAPR) worn by the subject while exercising in a neutral environment. A mathematical model of performance time as a function of the heat index (HI), respirator familiarity (RF), personality type (SN), and minute volume (Ve) indicated that performance time increased with a decrease in HI and Ve and with an increase in RF and SN ratio. A model of performance time as function of user acceptability and the heat index indicated that time to reach a level of "fairly uncomfortable" decreased exponentially from 5.34 minutes to 2.85 minutes with an increase in the heat index. Performance at the heat index conditions may be described by physiological, subjective, and individual characteristics.Item Development of a Transfer Function for Maximum Oxygen Deficit in Exercise While Wearing a Respiratory Protective Mask(2006-05-04) Phelps, Stephanie Jan; Johnson, Arthur T; Biological Resources Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Successful innovation of respirator masks depends on accurate models of exercise performance during respirator wear. Any valid model must include oxygen deficit (OD), which is a vital indicator of the physiological changes that occur during the transition from rest to exercise. OD represents anaerobic metabolism and is related to performance time. The goal of this research was to model the effect of a respirator on oxygen deficit. The following objectives were thereafter studied: (1) use experimental exercise data to calculate OD with a respirator; (2) determine the maximum OD and corresponding standard deviation values; and (3) develop a transfer function that accurately predicts OD in exercise while wearing a respirator. The study results indicated that oxygen deficit was significantly affected by exercise intensity and performance time; at 85% maximal capacity, respirator wear was not a significant factor affecting OD. Notably, the transfer function developed will serve a valuable predictive purpose.Item Texture-Based Segmentation and Finite Element Mesh Generation for Heterogeneous Biological Image Data(2005-04-12) Gudla, Prabhakar R; Montas, Hubert J; Biological Resources Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)The design, analysis, and control of bio-systems remain an engineering challenge. This is mainly due to the material heterogeneity, boundary irregularity, and nonlinear dynamics associated with these systems. The recent developments in imaging techniques and stochastic upscaling methods provides a window of opportunity to more accurately assess these bio-systems than ever before. However, the use of image data directly in upscaled stochastic framework can only be realized by the development of certain intermediate steps. The goal of the research presented in this dissertation is to develop a texture-segmentation method and a unstructured mesh generation for heterogeneous image data. The following two new techniques are described and evaluated in this dissertation: 1. A new texture-based segmentation method, using the stochastic continuum concepts and wavelet multi-resolution analysis, is developed for characterization of heterogeneous materials in image data. The feature descriptors are developed to efficiently capture the micro-scale heterogeneity of macro-scale entities. The materials are then segmented at a representative elementary scale at which the statistics of the feature descriptor stabilize. 2. A new unstructured mesh generation technique for image data is developed using a hierarchical data structure. This representation allows for generating quality guaranteed finite element meshes. The framework for both the methods presented in this dissertation, as such, allows them for extending to higher dimensions. The experimental results using these methods conclude them to be promising tools for unifying data processing concepts within the upscaled stochastic framework across biological systems. These are targeted for inclusion in decision support systems where biological image data, simulation techniques and artificial intelligence will be used conjunctively and uniformly to assess bio-system quality and design effective and appropriate treatments that restore system health.Item HYPERSPECTRAL IMAGING AND PATTERN RECOGNITION TECHNOLOGIES FOR REAL TIME FRUIT SAFETY AND QUALITY INSPECTION(2004-12-14) Cheng, Xuemei; Tao, Yang; Biological Resources Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Hyperspectral band selection and band combination has become a powerful tool and have gained enormous interest among researchers. An important task in hyperspectral data processing is to reduce the redundancy of the spectral and spatial information without losing any valuable details that are needed for the subsequent detection, discrimination and classification processes. An integrated principal component analysis (PCA) and Fisher linear discriminant (FLD) method has been developed for feature band selection, and other pattern recognition technologies have been applied and compared with the developed method. The results on different types of defects from cucumber and apple samples show that the integrated PCA-FLD method outperforms PCA, FLD and canonical discriminant methods when they are used separately for classification. The integrated method adds a new tool for the multivariate analysis of hyperspectral images and can be extended to other hyperspectral imaging applications. Dimensionality reduction not only serves as the first step of data processing that leads to a significant decrease in computational complexity in the successive procedures, but also a research tool for determining optimal spectra requirement for online automatic inspection of fruit. In this study, the hyperspectral research shows that the near infrared spectrum at 753nm is best for detecting apple defect. When applied for online apple defect inspection, over 98% of good apple detection rate is achieved. However, commercially available apple sorting and inspection machines cannot effectively solve the stem-calyx problems involved in automatic apple defects detection. In this study, a dual-spectrum NIR/MIR sensing method is applied. This technique can effectively distinguish true defects from stems and calyxes, which leads to a potential solution of the problem. The results of this study will advance the technology in fruit safety and quality inspection and improve the cost-effectiveness of fruit packing processes.Item DESIGN FOR A STAND-ALONE, UNIVERSAL SERIAL BUS (USB) -ENABLED AIRFLOW PERTURBATION DEVICE(2004-08-12) Silverman, Nischom K; Johnson, Arthur T; Biological Resources Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)The stand-alone, USB-enabled Airflow Perturbation Device (APD) provides an average respiratory resistance (RR) measurement without connection to a computer and offers expanded functionality when connected to a computer. In both home and medical clinic settings, RR can provide a measure of impairment in obstructive respiratory disorders and the effectiveness of respiratory therapies. The APD measures RR during passive breathing by sensing the ratio of pressure increase to flow reduction during brief, partial airflow interruptions. Prior work has shown the APD to produce repeatable, sensitive RR measurements in humans and animals. The device of prior investigations incorporated a computer and data acquisition card. The research presented here demonstrates that the APD can provide accurate measurements in a stand-alone format and provide expanded function with a USB host computer.Item MODEL OF EXERCISE PERFORMANCE WHILE WEARING A RESPIRATORY PROTECTIVE MASK(2004-06-29) Chiou, Yinghsiang; Johnson, Arthur T; Biological Resources Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)This model emphasized respiratory responses and incorporated mathematical descriptions of experimental results obtained from exercising humans. Prediction equations for tidal volume, anaerobic threshold, minute volume, respiratory work, and performance time were included, as well as dynamic changes in each. This model can help to design future respirators, aid workers wearing respirators, and regulate occupational health and safety. In general, the current model can predict performance time when subjects exercise both with and without masks. The current model was fitted for 30% and 80% VO2max of experimental data from the Human Performance Laboratory (University of Maryland, College Park). The results showed predicted values were reasonable and closer to the experimental data. Results of physiological values and performance times showed that the model structure was valid and that the model was capable of making rational predictions of the average effects of respirator wear on the pulmonary system during physical activity.