UBIR Repository (Univ. at Buffalo)
Not a member yet
5736 research outputs found
Sort by
Comparison of Bleaching Agent Applied to White Spot Lesions Before or After Resin Infiltration: An In Vitro Study
No full textM.S.Objectives: To compare the color alteration achieved from bleaching white spot lesions (WSLs) before resin infiltration (RI) relative to bleaching WSLs after RI, and to evaluate how these color alterations compare to adjacent areas of bleached sound enamel. Materials & Methods: Artificial WSLs were created on the right side of the facial surfaces of 90 extracted bovine maxillary incisors. A split-tooth design was performed with the left sides remaining as sound enamel. The specimens were randomly assigned into two groups (n = 45 per group), whereby WSLs were bleached then treated with resin infiltration (B-RI), or WSLs were treated with resin infiltration and then bleached (RI-B). Bleaching agent was applied over the entire tooth surface to incorporate both the left and right sides. For the left sides, color measurements (CIE L*, a*, b* values) were obtained at baseline and after bleaching procedures. For the right sides, color measurements were obtained at baseline, after WSL formation, and then after bleaching and RI procedures. Color measurements following bleaching and RI procedures were obtained after the specimens were subjected to a simulated 14-day thermocycling procedure. Inter- and intra-group color alterations (ΔE) were calculated at selected timepoints. The surfaces of representative specimens from each group were evaluated by scanning electron microscopy (SEM). Statistical analyses were performed using Mann-Whitney U and Wilcoxon signed-rank tests (P ≤0.05).Results: There were no statistically significant differences in color alteration (ΔE) when comparing WSLs that were bleached before or after RI. When determining the difference in color alterations (ΔE) between WSLs that were bleached then resin infiltrated, to the adjacent bleached sound enamel, there was no statistically significant difference. When determining the difference in color alterations (ΔE) between WSLs that were resin infiltrated and then bleached, to the adjacent bleached sound enamel, there was a statistically significant difference; however, the difference was deemed not to be of clinical significance. SEM analysis revealed that applying bleaching agent after RI mildly altered the surface texture of the resin material. Conclusions & Clinical Implications: The primary outcomes of this study indicate that color alterations are comparable whether bleach is applied to WSLs before or after RI. Therefore, if a patient bleaches after RI was previously completed, they can expect that their teeth will bleach homogeneously, despite RI restorations placed in areas of WSLs. The clinical implication of this study's findings is that it is best to abide by the common recommendation to bleach prior to RI therapy, in that although color alterations remain clinically comparable, SEM analysis revealed that applying bleaching agent after RI roughened the surface texture of the resin material.**To request an accessible version of the file(s) associated with this item, contact [email protected]. Please include the item's persistent URL [http://hdl.handle.net/. . .] in your request.*
Investigation of the Dosimetric Impact of Patient Immobilization Device and Treatment Couch Structures in Prone Breast Radiation Therapy
No full textPh.D.Purpose: In prone breast radiation, as the medial tangential beam usually passes through the immobilization board and couch, it is necessary to quantify the attenuation effect and the potential skin dose enhancement from these external structures when not included during the treatment planning process. Methods: The prone breast board studied consists of an insert on which the contralateral breast rests and a board base indexed to the couch. Two different Varian couch systems were also studied. The three effects being quantified were attenuation, scatter and bolus effect. Transmission factors (TF) of the board were measured using a Farmer chamber at 4 cm depth. Couch TFs were measured using a thimble chamber centered in a cylindrical phantom. TFs were then computed in the treatment planning system (TPS) for comparison. Selected clinical plans were recomputed in the TPS incorporating external structures for target coverage evaluation. Skin dose effects were measured using a Markus parallel plate chamber with a 1 mm buildup cap. A custom support model was created in the TPS for clinical implementation. The correction for the attenuation effect incorporating the custom support and couch structures in the treatment plans was also demonstrated. Monte Carlo (MC) simulations were carried out on phantom studies to further investigate the scatter and bolus effects from the breast board. Finally, the work flow of incorporating external structures in MC simulation for clinical prone breast cases were presented where an example of a patient case utilizing this workflow was carried out for a TPS-MC comparison. Results: Measured board insert / base TFs ranged 0.976-0.983 / 0.979-0.985 for 6MV and ranged 0.990-0.999 / 0.989-0.998 for 23MV x-rays, respectively, where TPS values agreed within 0.6%. Varian Exact Couch and Exact IGRT Couch TFs ranged 0.836-1 and 0.956-0.996, respectively, for 6 MV. The clinical treatment volume and whole breast receiving 95% of the prescription dose (CTV-V95 and WB-V95) of selected patients demonstrated reduced coverage due to attenuation of external structures. Close proximity to the base increased skin dose by up to 25-30%. Contacting the insert increased skin dose by 65-93% for 6MV and 117-157% for 23MV, respectively. For the custom support model of the breast board, assigned Hounsfield units (HUs) providing the best agreement was 200 and -100/-900 for the insert and board base respectively. Measured and MC values were compared for scatter effect with varying assigned densities (i.e. 0.25 g/cm3, 0.3 g/cm3, "sandwich structure") where the "sandwich structure" showed the best agreement. A dose enhancement was also seen in the MC simulation of the bolus effect from the insert. The MC dose calculation of an example clinical case showed a hot spot in an area near the board insert. Conclusion: Results have shown reduced coverage by attenuating external structures. Proper modeling of immobilization devices and couch structures in the TPS should be implemented for accurate dose calculation. Modelling a custom support structure in the TPS to match the measured attenuation results within a minimal margin of error is feasible. Increased surface doses were observed caused by direct contact to the insert or close proximity to the base. Further MC investigation with a sample of prone breast patients can potentially provide additional skin dose information for clinical decision.**To request an accessible version of the file(s) associated with this item, contact [email protected]. Please include the item's persistent URL [http://hdl.handle.net/. . .] in your request.*
A Comparative Analysis on Transformer Failure Modeling Techniques
No full textM.S.Power transformers are a vital part of electrical power utility systems. Like most electrical utility equipment under high stress conditions, power transformers are prone to failures and degradation in performance. A transformer failure can be detrimental to the utility network and be costly to replace. Producing a reliable model to predict transformer end-of-life is a topic that has been researched in the energy systems field. The main deterrent in these efforts has been the lack of detailed transformer failure data due to intellectual property rights, trade secrets, and the absence of operational data. The purpose of this work is to analyze modeling methods that aim to predict the time of failure for transformers based solely on age. Doble's Transformer Failure Subcommittee provided age-based data of 38,587 transformers for analysis. This data pool consisted of 35,712 operational units and 2,875 failed units. Each unit is a transmission-system transformer whose voltage ratings fall between 10kV and 700kV. The variable analyzed was the age of the transformer when the failure occurred. The age-of-failure data was analyzed using the Graphical Hazard Plotting Method to determine if an increased sample size would increase the modeling method's viability. However, additional parameters besides age influence transformer lifetimes. Diversity in voltage ratings, geographical locations, loading conditions and design characteristics vary between transformers. To address these additional parameters that were excluded in the initial model, alternative modeling methods including Health Index and Thermal Aging Modeling are examined during literature review. These methods typically include multiple failure mechanisms and modes that can contribute to transformer failures. The concept of hybrid modeling methods is introduced, and a proposed modeling structure is recommended. The proposed model incorporates the different failure mechanisms and modes that pertain to power transformers. It also accounts for historical failures and multi-factorial analysis to influence the health index scale. The hybrid structure can produce a more holistic model to predict transformer end-of-life. This method could enable utility companies with insights into failure-based maintenance schemes, targeted transformer testing, and spare/redundancy planning.**To request an accessible version of the file(s) associated with this item, contact [email protected]. Please include the item's persistent URL [http://hdl.handle.net/. . .] in your request.*
Native English Speaking Teachers' Classroom Discourse in Korean EFL Classrooms
No full textPh.D.Due to globalization and the 2015 Revised National Curriculum that set its goals to enhance learners' English communicative competence and understanding of the target culture, the number of native English speaking teachers (NESTs) teaching English in Korean public schools has been gradually increasing (Lee, 2015). In English as a Foreign Language (EFL) contexts, teachers serve as one of the main sources of language input to students, and, therefore, it is worthwhile to investigate their performance, including teachers' classroom discourse. This study contributes to this line of research. The subjects of the study consist of three NESTs and 137 students in novice and intermediate levels in a Korean university. Informed by socio-cultural theory (Mitchell & Myles, 2004) and the methodological framework of conversation analysis (Wooffitt, 2005), NESTs' classroom discourse in this study includes teachers' questioning strategies, co-construction of meaning with learners which is the scaffolding function of teacher talk, and corrective feedback. Based on one semester of class observations, interviews with teachers and students, and questionnaires with teachers and learners, the study found that Korean students tended not to respond when the teacher asked a question to the whole class because the students thought they might be wrong and wanted to save face in front of others. Accordingly, NESTs tried to give students more opportunities to speak in pairs or small groups, and they asked more focused questions to an individual student while walking around the classroom during group discussions. With respect to the questioning strategies, teachers asked more display questions than referential questions. However, in terms of referential questions, teachers asked more referential questions related to students' life-personal domains to an individual student and tended to stay away from asking those questions to the whole class. After teachers' questioning, when there was a communication breakdown, NESTs modified their discourse according to students for socio-cultural scaffolding by 1) using gesture or body language, 2) rewording or simplifying the utterance, 3) switching languages, 4) showing pictures and images online or drawing pictures, and 5) giving examples. Finally, as a follow up to students' response, teachers tended to give more one-on-one feedback instead of directing it to the whole class or in front of others so that students would not lose face or feel embarrassed. The current study contributes to an understudied area of research on NESTs' classroom discourse at the university level in Korea. It has implications for classroom pedagogy, including important issues of how questions and corrective feedback occur in EFL classrooms and how they are displayed in a culturally sensitive manner where English teachers try to find ways to maximize students' participation in class.**To request an accessible version of the file(s) associated with this item, contact [email protected]. Please include the item's persistent URL [http://hdl.handle.net/. . .] in your request.*
Fatigue Detection in Human-Robot Collaboration
No full textM.S.In this paper, a method is presented to detect muscle fatigue during Human-Robot Collaboration(HRC) using a low-cost myoelectric sensor. The proposed method uses the concepts of Riemannian geometry to extract simple and useful features from the data collected from muscle activity. A fine motor coordination task is designed and the human is asked to guide the robot along a virtual path. During the experiment, muscle activity from the dominant upper limb of the human is recorded. Then the subject is instructed to perform a muscle curl exercise until it induces fatigue in the muscles. Then the task is repeated with the fatigued muscles and the data is recorded. A total of nine subjects participated in the study and the experiments were labeled as before and after fatigue. At the end of the experiments, the muscle data from all nine subjects are used to extract Riemannian features. Previous studies in HRC have obtained good results by using other methods of extracting features from EMG data like obtaining wavelet coefficients by performing discrete wavelet transform and extracting time-domain features like RMS, MAV from the said coefficients, and Hudgin's set of time-domain features (RMS, MAV, WL, SSC, ZC) directly from the EMG data. These features are used to train classifiers to detect fatigue during human-robot collaboration. The accuracy and robustness of the classifier are compared against the conventional time-domain features and the time domain features extracted from the wavelets. We observe that the Riemannian features provide a better classification of the physical state of the human during an interaction. A classification accuracy of 92% is observed while using the Riemannian features against 86% accuracy with the conventional features and wavelet features. Moreover, other than the superior accuracy, the Riemannian features require fewer computations and can be calculated in real-time.**To request an accessible version of the file(s) associated with this item, contact [email protected]. Please include the item's persistent URL [http://hdl.handle.net/. . .] in your request.*
The Role of Trace Amine-Associated Receptor 1 in Nicotine Withdrawal
No full textM.S.Nicotine addiction remains a major health and social problem. Our previous study showed that trace amine-associated receptor 1(TAAR1), a target for the modulation of central dopaminergic activity, remarkably attenuates several abuse-related behaviors of nicotine in rats. However, whether TAAR1 plays a role in nicotine withdrawal remains unknown. Here, we examined the effects of TAAR1 in nicotine withdrawal using the partial agonist RO5263397. Rats underwent short-access (ShA) and long-access (LgA) nicotine self-administration training, respectively. Mecamylamine-precipitated withdrawal symptoms and elevated plus maze (EPM) test were performed to determine somatic withdrawal signs and anxiety-like behavior three days after the last training session. Von Frey filament tests were observed both at 1h and 72h after the last training session. We found that nicotine ShA rats achieved more infusions than the saline group, demonstrating the reinforcing effect of nicotine. Moreover, rats in LgA group obtained more nicotine than ShA group. ShA rats did not show significant withdrawal signs or anxiety-like behavior compared with saline group. In contrast, the LgA rats had more somatic withdrawal signs and spent more time in the open arm in the EPM test. More importantly, RO5263397 significantly reduced the somatic signs and anxiety-like behavior in LgA rats. There was no difference in the paw withdrawal threshold among all groups 1 hour after the training. However, LgA rats showed decreased paw withdrawal threshold compared to saline and ShA rats 72 hours after the last training, and RO5263397 completely reversed the hypersensitivity in LgA rats. These results strongly support that TAAR1 plays an important role in nicotine withdrawal.**To request an accessible version of the file(s) associated with this item, contact [email protected]. Please include the item's persistent URL [http://hdl.handle.net/. . .] in your request.*
Genotoxic Stress Causes Activation of Akt via Ca2+/Calmodulin-Dependent Protein Kinase Kinase 2 Pathway
No full textM.S.One of the major obstacles to the successful treatment of ovarian cancer is the rapid development of resistance to chemotherapy. Many genotoxic cancer treatments activate the serine/threonine kinase (Akt) pathway. Activation of this pathway plays a very important role in essential cellular functions such as survival, proliferation, migration, and angiogenesis that underlie the biology of human cancer. However, the mechanism by which Akt is activated in response to DNA damaging chemotherapeutic drugs and its effects on downstream targets are still unclear. In this study, we find that the treatment of ovarian cancer cells with the chemotherapeutic topoisomerase II inhibitor etoposide results in activating phosphorylation of Akt. A previous study from our laboratory reported that although the PI3K/ phosphoinositide-dependent kinase 1 (PDK1) pathway represents the canonical mechanism for phosphorylation of Akt at Thr-308, a non-canonical mechanism involves mediation of Akt phosphorylation at Thr-308 via Ca2+ /Calmodulin (CaM)-dependent protein kinase kinase 2 (CaMKK2). Since CaMKK2 is highly expressed in high-grade serous ovarian cancer, we investigated its role in Akt activation in ovarian cancer cells response to genotoxic stress caused by etoposide. We observed that Akt phosphorylation was significantly enhanced by etoposide in OVCAR-3 ovarian cancer cells and that RNAi-mediated knockdown of CaMKK2 significantly reduced the increase in activating Akt phosphorylation caused by etoposide. Consistent with an effect on p-Akt Thr-308 was an observed reduction in phosphorylation of the Akt downstream target PRAS40. Treatment of OVCAR3 cells with the intracellular calcium chelator, BAPTA-AM, or inhibition with the calmodulin inhibitor W-7 also decreased phosphorylation of Akt verifying that the action of CaMKK2 is reliant on its Ca2+/CaM-dependence.**To request an accessible version of the file(s) associated with this item, contact [email protected]. Please include the item's persistent URL [http://hdl.handle.net/. . .] in your request.*
Driving Safety of Autonomous Vehicles in Application Specific Environments
No full textM.S.In recent years, the widespread of autonomous vehicles unavoidably raises safety concerns for drivers, passengers and all road users alike. Advancement of autonomous vehicle safety technologies have come a long way. However, as the popularity of autonomous vehicles continues to gain traction, safety concerns must be addressed before the global application of autonomous vehicles, especially in more complex situations. This paper proposes an idea for a solution to ensure driving safety of autonomous vehicles in more complex situations. The solution is realized in the form of many software subsystems, subsequently tested and evaluated to study their theoretical application to real-world scenarios. The proposed solution aims to develop methods for autonomous vehicles to handle situations in regions of the world where road/traffic conditions are the most complex. Machine learning and computer vision are utilized, with the focus on developing non-Vehicle-to-Vehicle applications. We hope to foster further discussions and researches into suitable technologies and their applications, in regions of the world where complex situations may hinder the development and deployment of autonomous vehicles due to safety concerns.**To request an accessible version of the file(s) associated with this item, contact [email protected]. Please include the item's persistent URL [http://hdl.handle.net/. . .] in your request.*
High-Performance Computing and Data Science in Materials for Renewable Energy Science
No full textPh.D.Solar energy conversion to fuels via photo-electro-catalysis or to electricity via photovoltaics involves charge carriers (electrons and holes) that diffuse through the light harvesting material. Materials based on inexpensive and abundant transition metals are potentially great absorbers of solar light in the visible range. However, the main bottleneck to efficiency is slow charge carrier transport in these materials, a deficiency whose understanding and remediation requires atomistic level investigations. My research work is focused on the determination of carrier mobility in photo-active transition metal-based semiconductors and about finding correlations with their chemical compositions and structures. This work involves developing novel tools and approaches to characterize charge transfer in periodic crystalline systems...**To request an accessible version of the file(s) associated with this item, contact [email protected]. Please include the item's persistent URL [http://hdl.handle.net/. . .] in your request.*
The Roles of Word Frequency, Neighborhood Density, and Phonotactic Probability in Spoken Word Recognition in a Non-Native Language
No full textPh.D.Prior research has identified specific types of implicit linguistic knowledge that influence spoken word recognition (SWR), particularly word frequency, neighborhood density, and phonotactic probability. Almost all of this research has focused on native speakers, primarily English speakers. Much less research on these factors has examined their influence for non-native listeners. Research that does examine second or subsequent non-native language processing has focused on the effect of knowledge of sounds in a native language on the perception of non-native sounds (e.g., the influence of the knowledge of Japanese as a native language on the acquisition and recognition of English sounds r and l). Thus, little is known about how implicit linguistic knowledge in a non-native language affects word recognition in that language. The central goal of this study was to examine how higher-level or "top-down" language-specific knowledge, such as knowledge of word frequency, neighborhood density, and phonotactics, affects word recognition in a non-native language, and how experience with a non-native language modulates the use of such higher-level linguistic knowledge. The present study differed from previous studies of non-native SWR in several ways: (1) the effects of word frequency, neighborhood density, and phonotactic probability for native and non-native listeners of American English were examined; (2) stimuli consisted of verbs and were presented in a cross-modal identity priming paradigm in quiet (Experiment 1) and in a spoken word identification paradigm in noise (Experiment 2); (3) both accuracy (E1 and E2) and reaction time (E2) measures of processing were examined; (4) non-native listeners varied in their English language proficiency, which was measured on a continuous scale; (5) subjective word frequency ratings were collected to evaluate the efficacy of corpus derived frequency counts for non-native listeners; and (6) mixed-effects models were used to evaluate independent effects of the variables of interest while accounting for various control variables (e.g. trial order and prime duration) and for crossed random-effects for participants and items. The results revealed that native and non-native listeners were affected by the three factors of interest, indicating that non-native listeners had developed sensitivity to word frequency, neighborhood density, and phonotactic probability effects. In addition, English language proficiency had a large impact on performance. Less proficient non-native listeners generally displayed larger effects, reflecting less exposure to English word-forms and smaller lexicons, whereas more proficient listeners approached native listeners in their performance, reflecting more extensive experience in English and larger lexicons. Native listeners and more proficient non-native listeners displayed lexical competition effects. In contrast, less proficient non-native listeners appeared to show a segmental phonotactic probability effect, reflecting less experience in English but sensitivity to segmental phonotactic probabilities. Finally, interactions of the factors of interest with the verb type (regular, irregular, and irregular in the past tense) complicated the findings and may challenge the tacit assumption in the literature that word frequency, neighborhood density and phonotactic probability in SWR apply uniformly to all word types in all cases, irrespective of word morphology.**To request an accessible version of the file(s) associated with this item, contact [email protected]. Please include the item's persistent URL [http://hdl.handle.net/. . .] in your request.*