Esta tese propõe um método de avaliação de sistemas de obtenção de imagem radiológica através das Funções de Transferência simuladas para qualquer região do campo de radiação. Esse processo de simulação reflete as mudanças sofridas pelo ponto focal e, portanto, pela Função de Espalhamento de Ponto (FEP) ao longo do campo. A avaliação utiliza informações obtidas de Funções de Transferência ótica bi-dimensionais calculadas através da aplicação de Transformadas de Fourier sobre as Funções de Espalhamento de Ponto simuladas. O método permite realizar um mapeamento da resposta em freqüências espaciais dos sistemas radio1ógicos para todo o campo de radiação, comparar a nitidez da imagem de sistemas distintos com relação a regiões equivalentes e estudar uma região especifica para prever as distorções que serão inseridas na imagem. / In this work a method of evaluation for the radiological imaging systems performance by simulated transfer functions for any X-ray field region is proposed. The simulation procedures provide information of the changes of the focal spot sizes and, therefore, of the Point Spread Function (PSF) along the field. The evaluation yields the bi-dimensional optical Transfer Functions calculated from Fourier Transformation of the simulated PSFs. This method provides: the spatial frequency response for the entire radiation field of the radiological systems; comparisons of the image sharpness relative to equivalent field regions for different systems; and the study of a singular region in order to predict image distortions.

轉換公司債係指在一定條件下，能將該公司所發行的公司債，轉換為該公 司股票的金融債券，亦即轉換公司債兼具公司債和股票的雙重特性，因此 有必要對此一金融工具的評價方式進行了解。本研究之目的在於探討轉換 公司債之溢價理論，以及影響轉換公司債溢價之因素，並將其應用於臺灣 之轉換公司債，來分析其溢價行為，再建立轉換函數模型來估計與預測溢 價。本研究主要在於探討各個影響溢價因素對於溢價之影響，藉由整理 Brigham，Poensgen，Walter & Que，Weil、Segall & Green，Cretien ， Duvel，Mumey，West & Largay 等學者之溢價理論，再衡量臺灣之轉 換公司債市場，而選取股價變動性變數、轉換權利期間變數、未來所得差 異變數、價格底限變數以及交易成本差異變數五個變數，為迴歸模式中的 自變數，而溢價則為因變數。本研究之資料分析程序為：一、對於所選取 的五個自變數和一個因變數，分別建立單元迴歸，且進行逐步迴歸。二、 對於自變數和因變數建立複迴歸模型，利用刪除變數方法來解決線性重合 。三、將所得到無線性重合的自變數群和因變數，建立複迴歸模型，對其 進行t 檢定、 F檢定、自我相關檢定及殘差常態性檢定，若誤差項存在自 我相關，則建立時間數列方法中之轉換函數模型。四、利用轉換函數模型 將投入變數與產出變數，以一個動態體系相連結，經由轉換函數模型之認 定、估計、診斷性檢查之後，建立出一個最適模型，來對於溢價進行估計 與預測。本研究之研究對象為聲寶一及歌林一兩家轉換公司債，研究期間 為民國八十一年二月二十四日至民國八十三年五月一日，共114 週，而研 究結論為：一、聲寶一轉換公司債在對於溢價之單元迴歸中，轉換期間、 未來所得差異及價格底限三個變數，對溢價有顯著影響。通過線性重合檢 定的複迴歸模型中，只有股價變動性及未來所得差異，對於溢價的係數顯 著，且係數符號為正值。在轉換函數模型方面，投入變數（未來所得差異 變數）是以(1,0,0) 的形式影響溢價，且證明出轉換函數模型的預測力較 單變量模型佳。二、歌林一轉換公司債在對於溢價之單元迴歸中，股價變 動性、轉換期間、未來所得差異、價格底限及交易成本差異，這五個變數 ，對溢價均有顯著影響。通過線性重合檢定的複迴歸模型中，只有價格底 限變數，對於溢價的係數顯著，且係數符號為負值。。在轉換函數模型方 面，投入變數（價格底限變數）是以(0,2,0) 的形式影響溢價，且證明出 轉換函數模型的預測力較單變量模型佳。

在充滿不確定性之交易市場中，每位交易者會盡量利用所擁有之資訊，在 市場有干擾（如，風險性資產供給之不確定性、個人偏好不同、個人面對 之稅負環境不同等）之情形下，市場會顯露出部份私人訊息，故交易者亦 會經由對價格和交易量之觀察習得訊息；擁有私人訊息之交易者稱為消息 靈通者(Informed)，未擁有私人訊息而只能經由觀察價格而習(learn )得 訊息之交易者稱為消息不靈通者(Uninformed)，他們二者之差異在於他們 是否願花成本或資源以購買訊息。本文係在干擾理性預期模型下，利用所 設定之特殊效用函數--絕對風險規避效用函數及假設隨機變數為多元常態 分配，探討市場有干擾情形下，在第一期有私人訊息而在第二期有公開訊 息揭露之不對稱訊息模型中價格之資訊性，分別分析了公告訊息和私人訊 息之干擾程度、風險性資產供給之不確定及購買訊息人數對二期價格資訊 性之影響。在所設定的模型有解下，本文利用這些影響因素對公告訊息和 私人訊息在總合需求計劃部位 (Position)的彈性說明二期價格資訊性。 同時文中亦探討購買訊息人數之內生決定，顯示了公告訊息之揭露會修正 交易者之看法而減少私人蒐集訊息之誘因。

為因應預測目的不同，有時需要各種預測水平{\rm (forecast horizon)} ，例如，月預測可供進料、生產、補貨及倉儲之參考，年預測則可作為產 能規畫、產品線規畫、投資決策等之準則。然而，預測結構卻會因水平的 不同而彼此相異，以致產生諸多預測值的矛盾。有鑑於此，本研究主要以 一簡單且具理論基礎的整合{\rm intergration)} 過程，解決預測值互相 矛盾的問題。由於年資料通常屬於連續性模式，月資料則多為季節性模式 ，兩者透過的轉移函數形態截然不同，而且在解釋變數的選取上更是迥異 ，因此，需要經由加權平均的整合，才能使月預測值的加總等於年預測值 。至於權數的決定則以離散程度為準則，由於年資料為月資料的加總，兩 者均值相差甚多，故以變異係數為測量離散情形的標準。本研究主要乃遵 循{\rm Box-Jenkins} 的模式建立法則，構建連續性轉移函數模式及季節 性、轉移函數模式，並加以整合調整。在實證分析中以台灣啤酒銷售量為 例說明預測流程，年銷量預測方面以國民所得為解釋變數; 月銷量預測方 面則以氣溫為解釋變數，最後以加權平均將兩者整合調整。

近數十年來旅遊業不斷茁壯並走向多元化，在一國經濟中常扮演著舉足輕重地位。澳門為一小型開放經濟體，旅遊業是其產業龍頭及重點發展目標，對整體經濟發展至關重要。然回顧過去文獻少有考慮季節性與突發事件干擾的前提下，針對亞太地區的旅遊需求研究，加以預測方法多為單獨使用因果分析法或時間序列法。本研究在考慮該二種因素，先利用片段線性模型刻畫結構轉變的發生，再進一步結合時間序列分析法建構轉移函數模型。基於平均絕對比率誤差與均方差平方根的評比，轉移函數模型在預測澳門旅遊需求能力上確實有著較優異的表現。實證研究結果可供澳門旅遊相關業者規劃投資、觀光行銷企劃及政府單位採行總體政策參考。

The ongoing merger of the digital and optical components of the modern microscope is creating opportunities for new measurement techniques, along with new challenges for optical modelling. This thesis investigates several such opportunities and challenges which are particularly relevant to biomedical imaging. Fourier optics is used throughout the thesis as the underlying conceptual model, with a particular emphasis on three--dimensional Fourier optics. A new challenge for optical modelling provided by digital microscopy is the relaxation of traditional symmetry constraints on optical design. An extension of optical transfer function theory to deal with arbitrary lens pupil functions is presented in this thesis. This is used to chart the 3D vectorial structure of the spatial frequency spectrum of the intensity in the focal region of a high aperture lens when illuminated by linearly polarised beam. Wavefront coding has been used successfully in paraxial imaging systems to extend the depth of field. This is achieved by controlling the pupil phase with a cubic phase mask, and thereby balancing optical behaviour with digital processing. In this thesis I present a high aperture vectorial model for focusing with a cubic phase mask, and compare it with results calculated using the paraxial approximation. The effect of a refractive index change is also explored. High aperture measurements of the point spread function are reported, along with experimental confirmation of high aperture extended depth of field imaging of a biological specimen. Differential interference contrast is a popular method for imaging phase changes in otherwise transparent biological specimens. In this thesis I report on a new isotropic algorithm for retrieving the phase from differential interference contrast images of the phase gradient, using phase shifting, two directions of shear, and non--iterative Fourier phase integration incorporating a modified spiral phase transform. This method does not assume that the specimen has a constant amplitude. A simulation is presented which demonstrates good agreement between the retrieved phase and the phase of the simulated object, with excellent immunity to imaging noise.

The last few decades have been a very exciting period in thedevelopment of micro-electronics and brought us to the brink ofimplementing entire systems on a single chip, on a hithertounimagined scale. However an unforeseen challenge has croppedup in the form of managing wires, which have become the mainbottleneck in performance, masking the blinding speed of activedevices. A major problem is that increasingly complicatedeffects need to be modelled, but the computational complexityof any proposed model needs to be low enough to allow manyiterations in a design cycle. This thesis addresses the issue of closed form modelling ofthe response of coupled interconnect systems. Following astrict mathematical approach, second order models for thetransfer functions of coupled RC trees based on the first andsecond moments of the impulse response are developed. The2-pole-1-zero transfer function that is the best possible fromthe available information is obtained for the signal path fromeach driver to the output in multiple aggressor systems. Thisallows the complete response to be estimated accurately bysumming up the individual waveforms. The model represents theminimum complexity for a 2-pole-1-zero estimate, for this classof circuits. Also proposed are new techniques for the optimisation ofwires in on-chip buses. Rather than minimising the delay overeach individual wire, the configuration that maximises thetotal bandwidth over a number of parallel wires isinvestigated. It is shown from simulations that there is aunique optimal solution which does not necessarily translate tothe maximum possible number of wires, and in fact deviatesconsiderably from it when the resources available for repeatersare limited. Analytic guidelines dependent only on processparameters are derived for optimal sizing of wires andrepeaters. Finally regular tiled architectures with a commoncommunication backplane are being proposed as being the mostefficient way to implement systems-on-chip in the deepsubmicron regime. This thesis also considers the feasibility ofimplementing a regular packet-switched network-on-chip in atypical future deep submicron technology. All major physicalissues and challenges are discussed for two differentarchitectures and important limitations are identified.

The persistent thrust for a cleaner, greener environment has prompted air pollution regulations to be enforced with increased stringency by environmental protection bodies all over the world. This has prompted gas turbine manufacturers to move from non-premixed combustion to lean, premixed combustion. These lean premixed combustors operate quite fuel-lean compared to the stochiometric, in order to minimize CO and NOx productions, and are very susceptible to oscillations in any of the upstream flow variables. These oscillations cause the heat release rate of the flame to oscillate, which can engage one or more acoustic modes of the combustor or gas turbine components, and under certain conditions, lead to limit cycle oscillations. This phenomenon, called thermoacoustic instabilities, is characterized by very high pressure oscillations and increased heat fluxes at system walls, and can cause significant problems in the routine operability of these combustors, not to mention the occasional hardware damages that could occur, all of which cumulatively cost several millions of dollars. In a bid towards understanding this flow-flame interaction, this research works studies the heat release response of premixed flames to oscillations in reactant equivalence ratio, reactant velocity and pressure, under conditions where the flame preheat zone is convectively compact to these disturbances, using the G-equation. The heat release response is quantified by means of the flame transfer function and together with combustor acoustics, forms a critical component of the analytical models that can predict combustor dynamics. To this end, low excitation amplitude (linear) and high excitation amplitude (nonlinear) responses of the flame are studied in this work. The linear heat release response of lean, premixed flames are seen to be dominated by responses to velocity and equivalence ratio fluctuations at low frequencies, and to pressure fluctuations at high frequencies which are in the vicinity of typical screech frequencies in gas turbine combustors. The nonlinear response problem is exclusively studied in the case of equivalence ratio coupling. Various nonlinearity mechanisms are identified, amongst which the crossover mechanisms, viz., stoichiometric and flammability crossovers, are seen to be responsible in causing saturation in the overall heat release magnitude of the flame. The response physics remain the same across various preheat temperatures and reactant pressures. Finally, comparisons between the chemiluminescence transfer function obtained experimentally and the heat release transfer functions obtained from the reduced order model (ROM) are performed for lean, CH4/Air swirl-stabilized, axisymmetric V-flames. While the comparison between the phases of the experimental and theoretical transfer functions are encouraging, their magnitudes show disagreement at lower Strouhal number gains show disagreement.

<p>The last few decades have been a very exciting period in thedevelopment of micro-electronics and brought us to the brink ofimplementing entire systems on a single chip, on a hithertounimagined scale. However an unforeseen challenge has croppedup in the form of managing wires, which have become the mainbottleneck in performance, masking the blinding speed of activedevices. A major problem is that increasingly complicatedeffects need to be modelled, but the computational complexityof any proposed model needs to be low enough to allow manyiterations in a design cycle.</p><p>This thesis addresses the issue of closed form modelling ofthe response of coupled interconnect systems. Following astrict mathematical approach, second order models for thetransfer functions of coupled RC trees based on the first andsecond moments of the impulse response are developed. The2-pole-1-zero transfer function that is the best possible fromthe available information is obtained for the signal path fromeach driver to the output in multiple aggressor systems. Thisallows the complete response to be estimated accurately bysumming up the individual waveforms. The model represents theminimum complexity for a 2-pole-1-zero estimate, for this classof circuits.</p><p>Also proposed are new techniques for the optimisation ofwires in on-chip buses. Rather than minimising the delay overeach individual wire, the configuration that maximises thetotal bandwidth over a number of parallel wires isinvestigated. It is shown from simulations that there is aunique optimal solution which does not necessarily translate tothe maximum possible number of wires, and in fact deviatesconsiderably from it when the resources available for repeatersare limited. Analytic guidelines dependent only on processparameters are derived for optimal sizing of wires andrepeaters.</p><p>Finally regular tiled architectures with a commoncommunication backplane are being proposed as being the mostefficient way to implement systems-on-chip in the deepsubmicron regime. This thesis also considers the feasibility ofimplementing a regular packet-switched network-on-chip in atypical future deep submicron technology. All major physicalissues and challenges are discussed for two differentarchitectures and important limitations are identified.</p>

Direct Volume Rendering (DVR) is a visualization technique that has proved to be a very powerful tool in many scientific visualization applications. Diagnostic medical imaging is one domain where DVR could provide clear benefits in terms of unprecedented possibilities for analysis of complex cases and highly efficient work flow for certain routine examinations. The full potential of DVR in the clinical environment has not been reached, however, primarily due to limitations in conventional DVR methods and tools. This thesis presents methods addressing four major challenges for DVR in clinical use. The foundation of all methods is to incorporate the domain knowledge of the medical professional in the technical solutions. The first challenge is the very large data sets routinely produced in medical imaging today. To this end a multiresolution DVR pipeline is proposed, which dynamically prioritizes data according to the actual impact in the rendered image to be reviewed. Using this prioritization the system can reduce the data requirements throughout the pipeline and provide high performance and visual quality in any environment. Another problem addressed is how to achieve simple yet powerful interactive tissue classification in DVR. The methods presented define additional attributes that effectively captures readily available medical knowledge. The task of tissue detection is also important to solve in order to improve efficiency and consistency of diagnostic image review. Histogram-based techniques that exploit spatial relations in the data to achieve accurate and robust tissue detection are presented in this thesis. The final challenge is uncertainty visualization, which is very pertinent in clinical work for patient safety reasons. An animation method has been developed that automatically conveys feasible alternative renderings. The basis of this method is a probabilistic interpretation of the visualization parameters. Several clinically relevant evaluations of the developed techniques have been performed demonstrating their usefulness. Although there is a clear focus on DVR and medical imaging, most of the methods provide similar benefits also for other visualization techniques and application domains.