No description available.

Doctor of Philosophy / Radiographic film/screen (F/S) images have a narrow latitude or dynamic range. The film’s ability to record and view all the anatomy within the x-ray field is limited by this narrow dynamic range. The advent of digital radiographic means of storing and displaying radiographic images has improved the ability to record and visualise all of the anatomy. The problem still exists in digital radiography (DR) when radiographic examinations of certain anatomical regions are undertaken. In this work, the value of anatomically shaped radiographic contrast-enhancement masks (RCMs) in improving image contrast and reducing the dynamic range of images in DR was examined. Radiographic contrast-enhancement masks are digital masks that alter the radiographic contrast in DR images. The shape of these masks can be altered by the user. Anatomically shaped RCMs have been modelled on tissue compensation filters (TCFs) commonly used in F/S radiographic examinations. The prime purpose of a TCF is to reduce the dynamic range of photons reaching the image receptor and hence improve radiographic contrast in the resultant image. RCMs affect the dynamic range of the image rather than the energy source of the image, that of the x-ray photons. The research consisted of three distinct phases. The first phase was to examine physical TCFs and their effects on F/S radiographic images. Physical TCFs are used in radiographic F/S examinations to attenuate the x-ray beam to compensate for varying patient tissue thicknesses and/or densities. The effect of the TCF is to reduce resultant radiographic optical density variations in the image, allowing the viewer to observe a range of densities within the image which would otherwise not be visualised. Physical TCFs are commonly aluminium- or lead-based materials that attenuate the x-ray beam. A TCF has varying physical thickness to differentially attenuate the iii beam and is shaped for specific anatomical situations. During this project, various commonly used physical TCFs were examined. Measurements of size and thickness were made. Characteristics of linear attenuation coefficients and half-value thicknesses were delineated for various TCF materials and at various energies. The second phase of the research was to model the physical TCFs in a digital environment and apply the RCMs to DR images. The digital RCMs were created with similar characteristics to mimic the shapes to the physical TCFs. The RCM characteristics can be adjusted by the viewer of the image to suit the anatomy being imaged. Anatomically shaped RCMs were designed to assist in overcoming a limitation when viewing digital radiographic images, that of the dynamic range of the image. Anatomically shaped RCMs differ from other means of controlling the dynamic range of a digital radiographic image. It has been shown that RCMs can reduce the range of optical densities within images with a large dynamic range, to facilitate visualisation of all anatomy within the image. Physical TCFs are used within a specific range of radiographic F/S examinations. Digital radiographic images from this range of examinations were collected from various clinical radiological centres. Anatomically shaped RCMs were applied to the images to improve radiographic contrast of the images. The third phase of the research was to ascertain the benefits of the use of RCMs. Various other methods are currently in use to reduce the dynamic range of digital radiographic images. It is generally accepted that these methods also introduce noise into the image and hence reduce image quality. Quantitative comparisons of noise within the image were undertaken. The anatomically shaped RCMs introduced less noise than current methods designed to reduce the dynamic range of digital radiographic images. It was shown that RCM methods do not affect image quality. Radiographers make subjective assessment of digital radiographic image quality as part of their professional practice. To assess the subjective quality of images enhanced with anatomically shaped RCMs, a survey of radiographers and other iv qualified people was undertaken to ascertain any improvement in RCM-modified images compared to the original images. Participants were provided with eight pairs of image to compare. Questions were asked in the survey as to which image had the better range of optical densities; in which image the anatomy was easiest to visualise; which image had the simplest contrast and density manipulation for optimal visualisation; and which image had the overall highest image quality. Responses from 123 participants were received and analysed. The statistical analysis showed a higher preference by radiographers for the digital radiographic images in which the RCMs had been applied. Comparisons were made between anatomical regions and between patient-related factors of size, age and whether pathology was present in the image or not. The conclusion was drawn that digital RCMs correctly applied to digital radiographic images decrease the dynamic range of the image, allowing the entire anatomy to be visualised in one image. Radiographic contrast in the image can be maximised whilst maintaining image quality. Using RCMs in some digital radiographic examinations, radiographers will be able to present optimised images to referring clinicians. It is envisaged that correctly applied RCMs in certain radiographic examinations will enhance radiographic image quality and possibly lead to improved diagnosis from these images.

Doctor of Philosophy / Radiographic film/screen (F/S) images have a narrow latitude or dynamic range. The film’s ability to record and view all the anatomy within the x-ray field is limited by this narrow dynamic range. The advent of digital radiographic means of storing and displaying radiographic images has improved the ability to record and visualise all of the anatomy. The problem still exists in digital radiography (DR) when radiographic examinations of certain anatomical regions are undertaken. In this work, the value of anatomically shaped radiographic contrast-enhancement masks (RCMs) in improving image contrast and reducing the dynamic range of images in DR was examined. Radiographic contrast-enhancement masks are digital masks that alter the radiographic contrast in DR images. The shape of these masks can be altered by the user. Anatomically shaped RCMs have been modelled on tissue compensation filters (TCFs) commonly used in F/S radiographic examinations. The prime purpose of a TCF is to reduce the dynamic range of photons reaching the image receptor and hence improve radiographic contrast in the resultant image. RCMs affect the dynamic range of the image rather than the energy source of the image, that of the x-ray photons. The research consisted of three distinct phases. The first phase was to examine physical TCFs and their effects on F/S radiographic images. Physical TCFs are used in radiographic F/S examinations to attenuate the x-ray beam to compensate for varying patient tissue thicknesses and/or densities. The effect of the TCF is to reduce resultant radiographic optical density variations in the image, allowing the viewer to observe a range of densities within the image which would otherwise not be visualised. Physical TCFs are commonly aluminium- or lead-based materials that attenuate the x-ray beam. A TCF has varying physical thickness to differentially attenuate the iii beam and is shaped for specific anatomical situations. During this project, various commonly used physical TCFs were examined. Measurements of size and thickness were made. Characteristics of linear attenuation coefficients and half-value thicknesses were delineated for various TCF materials and at various energies. The second phase of the research was to model the physical TCFs in a digital environment and apply the RCMs to DR images. The digital RCMs were created with similar characteristics to mimic the shapes to the physical TCFs. The RCM characteristics can be adjusted by the viewer of the image to suit the anatomy being imaged. Anatomically shaped RCMs were designed to assist in overcoming a limitation when viewing digital radiographic images, that of the dynamic range of the image. Anatomically shaped RCMs differ from other means of controlling the dynamic range of a digital radiographic image. It has been shown that RCMs can reduce the range of optical densities within images with a large dynamic range, to facilitate visualisation of all anatomy within the image. Physical TCFs are used within a specific range of radiographic F/S examinations. Digital radiographic images from this range of examinations were collected from various clinical radiological centres. Anatomically shaped RCMs were applied to the images to improve radiographic contrast of the images. The third phase of the research was to ascertain the benefits of the use of RCMs. Various other methods are currently in use to reduce the dynamic range of digital radiographic images. It is generally accepted that these methods also introduce noise into the image and hence reduce image quality. Quantitative comparisons of noise within the image were undertaken. The anatomically shaped RCMs introduced less noise than current methods designed to reduce the dynamic range of digital radiographic images. It was shown that RCM methods do not affect image quality. Radiographers make subjective assessment of digital radiographic image quality as part of their professional practice. To assess the subjective quality of images enhanced with anatomically shaped RCMs, a survey of radiographers and other iv qualified people was undertaken to ascertain any improvement in RCM-modified images compared to the original images. Participants were provided with eight pairs of image to compare. Questions were asked in the survey as to which image had the better range of optical densities; in which image the anatomy was easiest to visualise; which image had the simplest contrast and density manipulation for optimal visualisation; and which image had the overall highest image quality. Responses from 123 participants were received and analysed. The statistical analysis showed a higher preference by radiographers for the digital radiographic images in which the RCMs had been applied. Comparisons were made between anatomical regions and between patient-related factors of size, age and whether pathology was present in the image or not. The conclusion was drawn that digital RCMs correctly applied to digital radiographic images decrease the dynamic range of the image, allowing the entire anatomy to be visualised in one image. Radiographic contrast in the image can be maximised whilst maintaining image quality. Using RCMs in some digital radiographic examinations, radiographers will be able to present optimised images to referring clinicians. It is envisaged that correctly applied RCMs in certain radiographic examinations will enhance radiographic image quality and possibly lead to improved diagnosis from these images.

Traumatic spinal cord injury (SCI) is a devastating condition for which there is no effective clinical treatment. Neuroprotective molecules that minimize tissue loss have shown promising results; however systemic delivery may limit in vivo benefits due to short systemic half-life and minimal passage across the blood-spinal cord barrier. To overcome these limitations, an injectable intrathecal delivery vehicle comprised of hyaluronan and methylcellulose (HAMC) was developed, and previously demonstrated to be safe and biocompatible intrathecally. Here, HAMC was determined to persist in the intrathecal space for between 4-7 d in vivo, indicating it as an optimal delivery system for neuroprotective agents to reduce tissue degeneration after SCI. HAMC was then investigated as an in vivo delivery system for two neuroprotective proteins: erythropoietin (EPO) and fibroblast growth factor 2 (FGF2). Both proteins demonstrated a diffusive release profile in vitro and maintained significant bioactivity during release. When EPO was delivered intrathecally with HAMC to the injured spinal cord, reduced cavitation in the tissue and significantly improved neuron counts were observed relative to the conventional delivery strategies of intraperitoneal and intrathecal bolus. When FGF2 was delivered intrathecally from HAMC, therapeutic concentrations penetrated into the injured spinal cord tissue for up to 6 h. Poly(ethylene glycol) modification of FGF2 significantly increased the amount of protein that diffused into the tissue when delivered similarly. Because FGF2 is a known angiogenic agent, dynamic computed tomography was developed for small animal serial assessment of spinal cord hemodynamics. Following SCI and treatment with FGF2 from HAMC, moderate improvement of spinal cord blood flow and a reduction in permeability were observed up to 7 d post-injury, suggesting that early delivery of neuroprotective agents can have lasting effects on tissue recovery. Importantly, the entirety of this work demonstrates that HAMC is an effective short-term delivery system for neuroprotective agents by improving tissue outcomes following traumatic SCI.

Traumatic spinal cord injury (SCI) is a devastating condition for which there is no effective clinical treatment. Neuroprotective molecules that minimize tissue loss have shown promising results; however systemic delivery may limit in vivo benefits due to short systemic half-life and minimal passage across the blood-spinal cord barrier. To overcome these limitations, an injectable intrathecal delivery vehicle comprised of hyaluronan and methylcellulose (HAMC) was developed, and previously demonstrated to be safe and biocompatible intrathecally. Here, HAMC was determined to persist in the intrathecal space for between 4-7 d in vivo, indicating it as an optimal delivery system for neuroprotective agents to reduce tissue degeneration after SCI. HAMC was then investigated as an in vivo delivery system for two neuroprotective proteins: erythropoietin (EPO) and fibroblast growth factor 2 (FGF2). Both proteins demonstrated a diffusive release profile in vitro and maintained significant bioactivity during release. When EPO was delivered intrathecally with HAMC to the injured spinal cord, reduced cavitation in the tissue and significantly improved neuron counts were observed relative to the conventional delivery strategies of intraperitoneal and intrathecal bolus. When FGF2 was delivered intrathecally from HAMC, therapeutic concentrations penetrated into the injured spinal cord tissue for up to 6 h. Poly(ethylene glycol) modification of FGF2 significantly increased the amount of protein that diffused into the tissue when delivered similarly. Because FGF2 is a known angiogenic agent, dynamic computed tomography was developed for small animal serial assessment of spinal cord hemodynamics. Following SCI and treatment with FGF2 from HAMC, moderate improvement of spinal cord blood flow and a reduction in permeability were observed up to 7 d post-injury, suggesting that early delivery of neuroprotective agents can have lasting effects on tissue recovery. Importantly, the entirety of this work demonstrates that HAMC is an effective short-term delivery system for neuroprotective agents by improving tissue outcomes following traumatic SCI.

Židininiai kepenų pakitimai (ŽKP) – vieni dažnesnių patologinių radinių tiriant pacientus dėl įvairių virškinimo ar kitų organizmo sistemų ligų. Įvairių spindulinės diagnostikos metodų dėka nustačius patologinius židinius kepenyse ypač svarbi jų klinikinė diferencinės diagnostikos reikšmė. ŽKP diferencinės diagnostikos esmė yra kiek įmanoma tikslesnis židinio kepenyse kraujotakos nustatymas ir aprašymas. Šios problemos sprendime naujų galimybių įgyja santykinai nauja ir Lietuvoje niekur kitur dar neatliekama UG tyrimo su i/v kontrastavimu metodika, kurios metu gaunami rezultatai lyginami su dominuojančių įprastų KT ir MRT kontrastinių tyrimų rezultatais. Tikslas: nustatyti radiologinių tyrimų vertę diagnozuojant židininius kepenų pakitimus (ŽKP) Uždaviniai: 1. Nustatyti įprastinio UG, UG su i/v kontrastavimu ir KT su i/v kontrastavimu diagnostikos metodų tikslumo rodiklius nepiktybiniams ir piktybiniams ŽKP. 2. Nustatyti UG ir KT su i/v kontrastavimu vertingiausius diagnostikos kriterijus vertinant židininių kepenų pokyčių prigimtį. 3. Palyginti tarpusavyje UG ir KT su i/v kontrastavimu nustatytų židininių kepenų pakitimų diagnostikos metodų tikslumo rodiklius nepiktybiniams ir piktybiniams ŽKP. 4. Įvertinti kepenų hemangiomų radiologinių diagnostikos metodų tikslumo vertę. / Focal hepatic lesions (FHL) are one of the most common pathological findings in patients who are examined for gastrointestinal or other diseases. The rigorous description of blood circulation in the hepatic focus is essential in differential diagnosis of FHL. US with i/v contrast enhancement is a new imaging method enabling definition of nature of FHL, follow-up, treatment efficacy and monitoring for possible relapse of the process. This study is conducted solely in the Radiology Department of Kaunas University of Medicine Hospital. Aim of the study To evaluate and compare the value of ‘bolus’ contrast-enhanced CT and contrast-enhanced US in diagnostics of focal hepatic lesions. Tasks of the study 1. To determine rates of diagnostic accuracy of conventional US, US with i/v contrast enhancement and ‘bolus’ contrast-enhanced CT in diagnosis of benign and malingnant FHL. 2. To establish the main and most effective diagnostic criteria of US and CT with i/v contrast enhancement for the assessment of nature of focal hepatic lesions. 3. To compare rates of diagnostic accuracy of US with i/v contrast enhancement and ‘bolus’ contrast-enhanced CT in diagnosis of benign and malingnant FHL. 4. To assess the reliability of radiological diagnostics of hemangioma, comparing the results of US, CT, and MRI.

With the advance of the LCD technology, video quality is becoming increasingly important. In this thesis, we develop hardware-friendly low-complexity enhancement algorithms. Video quality enhancement methods can be classified into two main categories. Single frame methods are the first category. These methods have generally low computational complexity. Multi-frame methods combine information from more than one frame and require the motion information of objects in the scene to do so. We first concentrate on the contrast-enhancement problem by using both global (frame-wise) and local information derived from the image. We use the image histogram and present a regularization-based histogram modification method to avoid problems that are often created by histogram equalization. Next, we design a compression artifact reduction algorithm that reduces ringing artifacts, which is disturbing especially on large displays. Furthermore, to remove the blurriness in the original video we present a non-iterative diffusion-based sharpening algorithm, which enhances edges in a ringing-aware fashion. The diffusion-based technique works on gradient approximations in a neighborhood individually. This gives more freedom compared to modulating the high-pass filter output that is used to sharpen the edges. Motion estimation enables applications such as motion-compensated noise reduction, frame-rate conversion, de-interlacing, compression, and super-resolution. Motion estimation is an ill-posed problem and therefore requires the use of prior knowledge on motion of objects. Objects have inertia and are usually larger then pixels or a block of pixels in size, which creates spatio-temporal correlation. We design a method that uses temporal redundancy to improve motion-vector search by choosing bias vectors from the previous frame and adaptively penalizes deviations from the bias vectors. This increases the robustness of the motion-vector search. The spatial correlation is more reliable because temporal correlation is difficult to use when the objects move fast or accelerate in time, or have small sizes. Spatial smoothness is not valid across motion boundaries. We investigate using energy minimization for motion estimation and incorporate the spatial smoothness prior into the energy. By formulating the energy minimization iterations for each motion vector as the primal problem, we show that the dual problem is motion segmentation for that specific motion vector.

As Deficiências na Percepção de Cores (DPC) afetam aproximadamente 200 milhões de pessoas em todo o mundo, comprometendo suas habilidades para efetivamente realizar tarefas relacionadas com cores e com visualização. Isto impacta significantemente os âmbitos pessoais e profissionais de suas vidas. Este trabalho apresenta um modelo baseado na fisiologia para simulação da percepção de cores. Além de modelar visão de cores normal, ele também compreende os tipos mais predominantes de deficiências na visão de cores (i.e., protanopia, deuteranopia, protanomalia e deuteranomalia), cujas causas são hereditárias. Juntos estes representam aproximadamente 99.96% de todos os casos de DPC. Para modelar a percepção de cores da visão humana, este modelo é baseado na teoria dos estágios e é derivado de dados reportados em estudos eletrofisiológicos. Ele é o primeiro modelo a consistentemente tratar visão de cores normal, tricromacia anômala e dicromacia de modo unificados. Seus resultados foram validados por avaliações experimentais envolvendo grupos de indivíduos com deficiência na percepção de cores e outros com visão de cores normal. Além disso, ele pode proporcionar a melhor compreensão e um feedback sobre como aperfeiçoar as experiências de visualização por indivíduos com DPC. Ele também proporciona um framework para se testar hipóteses sobre alguns aspectos acerca das células fotoreceptoras na retina de indivíduos com deficiência na percepção de cores. Este trabalho também apresenta uma técnica automática de recoloração de imagens que visa realçar o contraste de cores para indivíduos dicromatas com custo computacional variando linearmente com o número de pixels. O algoritmo proposto pode ser eficientemente implementado em GPUs, e para imagens com tamanhos tipicos ele apresenta performance de até duas ordens de magnitude mais rápida do que as técnicas estado da arte atuais. Ao contrário das abordagens anteriores, a técnica proposta preserva coerência temporal e, portanto, é adequado para recoloração de vídeos. Este trabalho demonstra a efetividade da técnica proposta ao integrá-la a um sistema de visualização e apresentando, pela primeira vez, cenas de visualização recoloridas para dicromatas em tempo-real e com alta qualidade. / Color vision deficiency (CVD) affects approximately 200 million people worldwide, compromising the ability of these individuals to effectively perform color and visualizationrelated tasks. This has a significant impact on their private and professional lives. This thesis presents a physiologically-based model for simulating color perception. Besides modeling normal color vision, it also accounts for the hereditary and most prevalent cases of color vision deficiency (i.e., protanopia, deuteranopia, protanomaly, and deuteranomaly), which together account for approximately 99.96% of all CVD cases. This model is based on the stage theory of human color vision and is derived from data reported in electrophysiological studies. It is the first model to consistently handle normal color vision, anomalous trichromacy, and dichromacy in a unified way. The proposed model was validated through an experimental evaluation involving groups of color vision deficient individuals and normal color vision ones. This model can provide insights and feedback on how to improve visualization experiences for individuals with CVD. It also provides a framework for testing hypotheses about some aspects of the retinal photoreceptors in color vision deficient individuals. This thesis also presents an automatic image-recoloring technique for enhancing color contrast for dichromats whose computational cost varies linearly with the number of input pixels. This approach can be efficiently implemented on GPUs, and for typical image sizes it is up to two orders of magnitude faster than the current state-of-the-art technique. Unlike previous approaches, the proposed technique preserves temporal coherence and, therefore, is suitable for video recoloring. This thesis demonstrates the effectiveness of the proposed technique by integrating it into a visualization system and showing, for the first time, real-time high-quality recolored visualizations for dichromats.

As Deficiências na Percepção de Cores (DPC) afetam aproximadamente 200 milhões de pessoas em todo o mundo, comprometendo suas habilidades para efetivamente realizar tarefas relacionadas com cores e com visualização. Isto impacta significantemente os âmbitos pessoais e profissionais de suas vidas. Este trabalho apresenta um modelo baseado na fisiologia para simulação da percepção de cores. Além de modelar visão de cores normal, ele também compreende os tipos mais predominantes de deficiências na visão de cores (i.e., protanopia, deuteranopia, protanomalia e deuteranomalia), cujas causas são hereditárias. Juntos estes representam aproximadamente 99.96% de todos os casos de DPC. Para modelar a percepção de cores da visão humana, este modelo é baseado na teoria dos estágios e é derivado de dados reportados em estudos eletrofisiológicos. Ele é o primeiro modelo a consistentemente tratar visão de cores normal, tricromacia anômala e dicromacia de modo unificados. Seus resultados foram validados por avaliações experimentais envolvendo grupos de indivíduos com deficiência na percepção de cores e outros com visão de cores normal. Além disso, ele pode proporcionar a melhor compreensão e um feedback sobre como aperfeiçoar as experiências de visualização por indivíduos com DPC. Ele também proporciona um framework para se testar hipóteses sobre alguns aspectos acerca das células fotoreceptoras na retina de indivíduos com deficiência na percepção de cores. Este trabalho também apresenta uma técnica automática de recoloração de imagens que visa realçar o contraste de cores para indivíduos dicromatas com custo computacional variando linearmente com o número de pixels. O algoritmo proposto pode ser eficientemente implementado em GPUs, e para imagens com tamanhos tipicos ele apresenta performance de até duas ordens de magnitude mais rápida do que as técnicas estado da arte atuais. Ao contrário das abordagens anteriores, a técnica proposta preserva coerência temporal e, portanto, é adequado para recoloração de vídeos. Este trabalho demonstra a efetividade da técnica proposta ao integrá-la a um sistema de visualização e apresentando, pela primeira vez, cenas de visualização recoloridas para dicromatas em tempo-real e com alta qualidade. / Color vision deficiency (CVD) affects approximately 200 million people worldwide, compromising the ability of these individuals to effectively perform color and visualizationrelated tasks. This has a significant impact on their private and professional lives. This thesis presents a physiologically-based model for simulating color perception. Besides modeling normal color vision, it also accounts for the hereditary and most prevalent cases of color vision deficiency (i.e., protanopia, deuteranopia, protanomaly, and deuteranomaly), which together account for approximately 99.96% of all CVD cases. This model is based on the stage theory of human color vision and is derived from data reported in electrophysiological studies. It is the first model to consistently handle normal color vision, anomalous trichromacy, and dichromacy in a unified way. The proposed model was validated through an experimental evaluation involving groups of color vision deficient individuals and normal color vision ones. This model can provide insights and feedback on how to improve visualization experiences for individuals with CVD. It also provides a framework for testing hypotheses about some aspects of the retinal photoreceptors in color vision deficient individuals. This thesis also presents an automatic image-recoloring technique for enhancing color contrast for dichromats whose computational cost varies linearly with the number of input pixels. This approach can be efficiently implemented on GPUs, and for typical image sizes it is up to two orders of magnitude faster than the current state-of-the-art technique. Unlike previous approaches, the proposed technique preserves temporal coherence and, therefore, is suitable for video recoloring. This thesis demonstrates the effectiveness of the proposed technique by integrating it into a visualization system and showing, for the first time, real-time high-quality recolored visualizations for dichromats.

As Deficiências na Percepção de Cores (DPC) afetam aproximadamente 200 milhões de pessoas em todo o mundo, comprometendo suas habilidades para efetivamente realizar tarefas relacionadas com cores e com visualização. Isto impacta significantemente os âmbitos pessoais e profissionais de suas vidas. Este trabalho apresenta um modelo baseado na fisiologia para simulação da percepção de cores. Além de modelar visão de cores normal, ele também compreende os tipos mais predominantes de deficiências na visão de cores (i.e., protanopia, deuteranopia, protanomalia e deuteranomalia), cujas causas são hereditárias. Juntos estes representam aproximadamente 99.96% de todos os casos de DPC. Para modelar a percepção de cores da visão humana, este modelo é baseado na teoria dos estágios e é derivado de dados reportados em estudos eletrofisiológicos. Ele é o primeiro modelo a consistentemente tratar visão de cores normal, tricromacia anômala e dicromacia de modo unificados. Seus resultados foram validados por avaliações experimentais envolvendo grupos de indivíduos com deficiência na percepção de cores e outros com visão de cores normal. Além disso, ele pode proporcionar a melhor compreensão e um feedback sobre como aperfeiçoar as experiências de visualização por indivíduos com DPC. Ele também proporciona um framework para se testar hipóteses sobre alguns aspectos acerca das células fotoreceptoras na retina de indivíduos com deficiência na percepção de cores. Este trabalho também apresenta uma técnica automática de recoloração de imagens que visa realçar o contraste de cores para indivíduos dicromatas com custo computacional variando linearmente com o número de pixels. O algoritmo proposto pode ser eficientemente implementado em GPUs, e para imagens com tamanhos tipicos ele apresenta performance de até duas ordens de magnitude mais rápida do que as técnicas estado da arte atuais. Ao contrário das abordagens anteriores, a técnica proposta preserva coerência temporal e, portanto, é adequado para recoloração de vídeos. Este trabalho demonstra a efetividade da técnica proposta ao integrá-la a um sistema de visualização e apresentando, pela primeira vez, cenas de visualização recoloridas para dicromatas em tempo-real e com alta qualidade. / Color vision deficiency (CVD) affects approximately 200 million people worldwide, compromising the ability of these individuals to effectively perform color and visualizationrelated tasks. This has a significant impact on their private and professional lives. This thesis presents a physiologically-based model for simulating color perception. Besides modeling normal color vision, it also accounts for the hereditary and most prevalent cases of color vision deficiency (i.e., protanopia, deuteranopia, protanomaly, and deuteranomaly), which together account for approximately 99.96% of all CVD cases. This model is based on the stage theory of human color vision and is derived from data reported in electrophysiological studies. It is the first model to consistently handle normal color vision, anomalous trichromacy, and dichromacy in a unified way. The proposed model was validated through an experimental evaluation involving groups of color vision deficient individuals and normal color vision ones. This model can provide insights and feedback on how to improve visualization experiences for individuals with CVD. It also provides a framework for testing hypotheses about some aspects of the retinal photoreceptors in color vision deficient individuals. This thesis also presents an automatic image-recoloring technique for enhancing color contrast for dichromats whose computational cost varies linearly with the number of input pixels. This approach can be efficiently implemented on GPUs, and for typical image sizes it is up to two orders of magnitude faster than the current state-of-the-art technique. Unlike previous approaches, the proposed technique preserves temporal coherence and, therefore, is suitable for video recoloring. This thesis demonstrates the effectiveness of the proposed technique by integrating it into a visualization system and showing, for the first time, real-time high-quality recolored visualizations for dichromats.