Loading...
Publication
Estudo das propriedades óticas espetrais de tecidos do córtex cerebral de coelho
| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 4.96 MB | Adobe PDF |
Authors
Advisor(s)
Abstract(s)
O cérebro é o órgão mais complexo no corpo humano e aquele que nos permite aprender a um ritmo elevado. Dada a sua complexidade interna e a função importante que desempenha, o cérebro é talvez de todos os órgãos o menos explorado e menos estudado. No campo da ótica médica vários estudos foram já realizados com tecidos cerebrais, fornecendo cada um desses estudos informação fisiológica e de diagnóstico de elevada importância. No presente trabalho foram medidos espectros entre 200 e 1000 nm, com diferentes montagens para obter as propriedades óticas espectrais que caracterizam o córtex cerebral do coelho. Após determinado o conjunto fundamental das propriedades óticas que caracterizam o córtex cerebral, uma análise mais cuidada do coeficiente de absorção mostrou que a existência de melanina e de lipofuscina no córtex ocultava os reais conteúdos de ADN e de hemoglobina neste tecido. Construindo o contributo de absorção destes dois pigmentos verificou-se que o córtex continha quase o dobro de melanina em relação à lipofuscina, o que se veio a associar a processos degenerativos nos neurónios do córtex. Subtraindo o contributo de absorção destes pigmentos ao coeficiente de absorção, tornou possível a avaliação do real conteúdo de ADN e de hemoglobina no córtex, tendo-se obtido valores da mesma ordem de grandeza aos observados para outros tecidos biológicos. Numa tentativa de verificar se tais resultados podiam ser reproduzidos com medições feitas numa perspetiva não invasiva, foram aplicadas técnicas de machine learning a espectros medidos com uma montagem de reflectância difusa. Os resultados obtidos nestas estimações mostraram limitações relacionadas com o baixo número de espectros que foram medidos. No entanto, tais resultados mostraram-se próximos dos resultados obtidos por cálculo direto a partir de medições invasivas e com características fisiológicas e de diagnóstico semelhantes. O conteúdo de melanina superior no córtex permaneceu nos resultados estimados, tal como os conteúdos de ADN e de hemoglobina. Tais resultados mostram que apesar de se ter usado um baixo número de espectros para desenvolver os algoritmos de estimação do espectro do coeficiente de absorção, estes são capazes de detetar as características fisiológicas de interesse no tecido. O procedimento usado em todo o estudo, mas com um maior número de espectros para estimar o coeficiente de absorção de forma não invasiva, poderá ser usado noutros tecidos saudáveis e com patologias para desenvolver novas técnicas de diagnóstico.
The brain is the most complex organ in the human body and the one that allows us to learn at a high rate. Due to its internal complexity and the important function that it plays, the brain is perhaps the less explored and less studied among the organs of the body. Regarding Biophotonics, several studies have previously been made with brain tissues, providing physiological and diagnostic information of great importance. In the present study, spectra were measured between 200 and 1000 nm, using with different setups to obtain the spectral optical properties of the brain cortex of rabbit. Once the fundamental set of optical properties that characterizes the brain cortex was obtained, a careful analysis of the absorption coefficient showed that the presence of melanin and lipofuscin in the cortex was hiding the real contents of DNA and haemoglobin in this tissue. Constructing the absorption contribution of these two pigments, it was verified that the cortex had almost the double content of melanin in regard to the lipofuscin content, a result that was associated to neurodegenerative processes in the cortex. Subtracting the absorption contribution of these pigments from the absorption coefficient allowed the evaluation of the real DNA and haemoglobin contents in the cortex, showing values in the same order as the ones observed for other biological tissues. In an attempt to check if those results could be reproduced with measurements made with a noninvasive approach, machine learning techniques were applied to diffuse reflectance spectra. The results of those estimations showed some limitations that are directly related to the reduced number of measured spectra. Nevertheless, such estimated results showed to be close to the ones obtained through direct calculation from spectral measurements that were made in an invasive procedure. The estimated results were also able to show similar physiological and diagnostic characteristics as the ones obtained from direct calculations. The higher melanin content remained in the estimated results, as well as the DNA and haemoglobin contents. Such results show that even though a low number of spectra was used to develop the estimation algorithms to obtain the absorption coefficient, the estimated results also contain the interesting physiological characteristics for this tissue. The procedure used in the present study, but considering a higher number of spectra to perform the estimation of the absorption coefficient from noninvasive measurements can be used with other healthy and pathological tissues, so that new diagnostic techniques can be developed.
The brain is the most complex organ in the human body and the one that allows us to learn at a high rate. Due to its internal complexity and the important function that it plays, the brain is perhaps the less explored and less studied among the organs of the body. Regarding Biophotonics, several studies have previously been made with brain tissues, providing physiological and diagnostic information of great importance. In the present study, spectra were measured between 200 and 1000 nm, using with different setups to obtain the spectral optical properties of the brain cortex of rabbit. Once the fundamental set of optical properties that characterizes the brain cortex was obtained, a careful analysis of the absorption coefficient showed that the presence of melanin and lipofuscin in the cortex was hiding the real contents of DNA and haemoglobin in this tissue. Constructing the absorption contribution of these two pigments, it was verified that the cortex had almost the double content of melanin in regard to the lipofuscin content, a result that was associated to neurodegenerative processes in the cortex. Subtracting the absorption contribution of these pigments from the absorption coefficient allowed the evaluation of the real DNA and haemoglobin contents in the cortex, showing values in the same order as the ones observed for other biological tissues. In an attempt to check if those results could be reproduced with measurements made with a noninvasive approach, machine learning techniques were applied to diffuse reflectance spectra. The results of those estimations showed some limitations that are directly related to the reduced number of measured spectra. Nevertheless, such estimated results showed to be close to the ones obtained through direct calculation from spectral measurements that were made in an invasive procedure. The estimated results were also able to show similar physiological and diagnostic characteristics as the ones obtained from direct calculations. The higher melanin content remained in the estimated results, as well as the DNA and haemoglobin contents. Such results show that even though a low number of spectra was used to develop the estimation algorithms to obtain the absorption coefficient, the estimated results also contain the interesting physiological characteristics for this tissue. The procedure used in the present study, but considering a higher number of spectra to perform the estimation of the absorption coefficient from noninvasive measurements can be used with other healthy and pathological tissues, so that new diagnostic techniques can be developed.
Description
Keywords
Córtex cerebral Índice de refração Coeficiente de absorção Coeficiente de espalhamento Coeficiente de espalhamento reduzido Anisotropia de espalhamento Profundidade de penetração da luz Espectros Melanina Lipofuscina Hemoglobina ADN Machine learning Processos neuro degenerativos Derrame cerebral Alzheimer Parkinson Brain cortex Refractive index Absorption coefficient Scattering coefficient Reduced scattering coefficient Scattering anisotropy Light penetration depth Spectra Melanin Lipofuscin Haemoglobin Neurodegenerative processes Brain stroke