| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 1.91 MB | Adobe PDF |
Authors
Abstract(s)
A presente dissertação prende-se pelo processamento e análise computacional de cintigrafias de Glândulas Salivares em pacientes com suspeita de Síndrome de Sjogren. Com esse propósito foram implementados algoritmos computacionais para segmentação e análise de imagens para aplicação em estruturas como as glândulas salivares e a glândula tiróide. O objetivo de avaliar a resposta à terapêutica em pacientes com suspeita de síndrome de Sjorgen advém da necessidade de avaliar quantitativamente a resposta à
terapêutica prescrita. Normalmente os pacientes com este tipo de patologia realizam cintigrafias das glândulas salivares periódicas, que se limitam avaliar a progressão ou regressão da doença. No entanto, esta análise é meramente qualitativa. Posto isto, foi desenvolvida uma metodologia para segmentar manualmente as regiões referidas através de imagens obtidas em Medicina Nuclear, mais especificamente cintigrafias das glândulas salivares com administração do radiofármaco Na99mTcO4. Os modelos desenvolvidos permitiram segmentar as regiões de interesse e aplicar esta segmentação em novas imagens. À vista disso, a utilização de tais técnicas proporcionou a extração de medidas quantitativas nessas regiões que foram utilizadas para determinar os índices de captação de radiofármaco em focos hipercaptantes,
nomeadamente as glândulas salivares e a glândula tiróide para o despiste de Síndrome de Sjorgen.
Os resultados obtidos pela aplicação da metodologia indicam que a eficácia do modelo depende vigorosamente da imagem e da informação que está a ser extraída das regiões. De um ponto de vista global, o software desenvolvido produziu resultados satisfatórios, no entanto, a longo prazo, o algoritmo possa ser melhorado e alargado a um maior número de exames. Em suma, o modelo desenvolvido permite avaliar a terapêutica em pacientes com suspeita de Síndrome de Sjogren.
The present dissertation was performed using computational analysis of salivary gland scintigraphy in patients with suspected Sjogren's Syndrome. With this topic we have implemented computational algorithms for segmentation and image analysis for application in structures such as salivary glands and thyroid gland. The objective of assessing the response to therapy in patients with suspected Sjorgen syndrome arises from the need to quantitatively assess response to prescribed therapy. Usually patients with this type of pathology perform scintigraphy of the periodic salivary glands, which are limited to evaluate the progression or regression of the disease. However, this analysis is merely qualitative. Therefore, a methodology was developed to segment the referred regions through images obtained in Nuclear Medicine, more specifically salivary gland scans with administration of the radiopharmaceutical Na99mTcO4. The models developed allowed us to segment the regions of interest and apply this segmentation to new images. In view of this, the use of such techniques provided the extraction of quantitative measurements in these regions that were used to determine the rates of uptake of radiopharmaceuticals in hypercapitating foci, namely the salivary glands and thyroid gland for the diagnosis of Sjorgen's Syndrome. The results obtained by applying the methodology indicate that the efficacy of the model depends strongly on the image and the information that is being extracted from the regions. From a global point of view, the developed software produced satisfactory results, however, in the long run, the algorithm can be improved and extended to a greater number of exams. In summary, the developed model allows to evaluate the therapy in patients with suspected Sjogren Syndrome.
The present dissertation was performed using computational analysis of salivary gland scintigraphy in patients with suspected Sjogren's Syndrome. With this topic we have implemented computational algorithms for segmentation and image analysis for application in structures such as salivary glands and thyroid gland. The objective of assessing the response to therapy in patients with suspected Sjorgen syndrome arises from the need to quantitatively assess response to prescribed therapy. Usually patients with this type of pathology perform scintigraphy of the periodic salivary glands, which are limited to evaluate the progression or regression of the disease. However, this analysis is merely qualitative. Therefore, a methodology was developed to segment the referred regions through images obtained in Nuclear Medicine, more specifically salivary gland scans with administration of the radiopharmaceutical Na99mTcO4. The models developed allowed us to segment the regions of interest and apply this segmentation to new images. In view of this, the use of such techniques provided the extraction of quantitative measurements in these regions that were used to determine the rates of uptake of radiopharmaceuticals in hypercapitating foci, namely the salivary glands and thyroid gland for the diagnosis of Sjorgen's Syndrome. The results obtained by applying the methodology indicate that the efficacy of the model depends strongly on the image and the information that is being extracted from the regions. From a global point of view, the developed software produced satisfactory results, however, in the long run, the algorithm can be improved and extended to a greater number of exams. In summary, the developed model allows to evaluate the therapy in patients with suspected Sjogren Syndrome.
Description
Keywords
Glândulas salivares Processamento e análise de imagem Síndrome de Sjogren Segmentação de objeto Salivary glands Image processing and analysis Sjogren's Syndrome Object segmentation