Método de avaliação das curvaturas sagitais da coluna vertebral em um ambiente computacional

Abstract

Metodologias não invasivas de avaliação postural avaliam a coluna vertebral a partir da superfície da pele na tentativa de estimar a curvatura interna da coluna. Porém, alguns erros são encontrados no que diz respeito aos pontos anatômicos utilizados por essas metodologias. Objetivo: desenvolver, em um ambiente computacional, uma metodologia para mensurar as curvaturas sagitais (torácica e lombar) externas da coluna vertebral, a partir da superfície da pele, que seja concordante com as curvaturas internas sagitais da coluna vertebral. Metodologia: Para o desenvolvimento das rotinas de avaliação das magnitudes da coluna vertebral, foram utilizados 15 Raios-X digitais panorâmicos da coluna vertebral de crianças voluntárias provenientes de uma base de dados pertencente ao Grupo de Pesquisa Biomec/UFRGS. Foram marcados nas imagens os seguintes pontos anatômicos de referência: quatro vértices do corpo vertebral (ântero-superior, póstero-superior, ântero-inferior e póstero-inferior), ápice do processo espinhoso e ponto da superfície da pele referente à localização do processo espinhoso. Todos esses pontos foram digitalizados para as seguintes vértebras: C6, C7, T2, T4, T6, T8, T10, T12, L2, L4 e S2. A rotina de avaliação foi desenvolvida na linguagem de programação utilizada no software Matlab da Mathworks. Para a cifose torácica e lordose lombar foram obtidas três curvaturas através de um polinômio de 3ª ordem: (1) curvatura interna a partir do baricentro dos corpos vertebrais; (2) curvatura interna a partir do ápice dos processos espinhosos; (3) curvatura externa a partir da superfície da pele. Para o cálculo do ângulo de cada curvatura da cifose torácica e da lordose lombar foram obtidas as tangentes relativas aos pontos de interesse.

Non-invasive assessment methodologies for postural evaluation of the spine from the surface of the skin in an attempt to estimate the internal curvature of the spine. However, some errors are encountered with respect to anatomical landmarks used by these methodologies. Objective: develop, in a computing environment, a method for measuring sagittal curvatures (thoracic and lumbar) outside the spinal column, from the skin surface, which is consistent with the internal sagittal curvature of the spine. Methodology: For the development of routine assessment of the spine magnitudes, 15 digital panoramic X-ray of the spine of volunteer children were used from a database belonging to Biomec Research Group / UFRGS. The following anatomical landmarks were marked on the images: four vertices of the vertebral body (anterior superior, posterior superior, anterior inferior and posterior inferior), apex of the spinous process and a point on the skin surface concerning the location of the spinous process. All these points were digitalized for the following vertebrae, C6, C7, T2, T4, T6, T8, T10, T12, L2, L4 and S2. The evaluation routine was developed in the programming language used in Matlab software from Mathworks. For thoracic kyphosis and lumbar lordosis three curvatures were obtained through a 3rd order polynomial: (1) internal curvature from the centroid of the vertebral bodies; (2) internal curvature from the apex of the spinous processes; (3) External curvature from the surface of the skin. To calculate the angle of each bend of thoracic kyphosis and lumbar lordosis, tangents concerning points of interest were obtained. The vertebral levels used for kyphosis were: C7-T12, C7-T10, T2-T12, T2-T10, T4-T12 and T4-T10; and for lordosis were: T10-S2, T10-L4, T12-S2 e T12-L4. For statistical analysis the Product Moment Correlation Test and the Pearson Test of Simple Linear Regression were used. α=0,05. Results: For thoracic kyphosis, vertebral levels that had the highest correlation coefficients among the three curvatures were: T4-T12 (r between 0,795-0,942; p<0,05) and T2-T12 (r between 0,679-0,921; p<0,05). For lumbar lordosis, vertebral levels T10-S2 (r between 0.887 to 0.925; p <0.05) and T12-S2 (r between 0,836-0,927; p<0,05) showed the best correlation among the three curves. Through the line equation obtained by the linear regression, the angles provided by the skin surface can approach the angles obtained from the internal bends, making the surface evaluation more accurate in relation to the internal evaluation of the spine. Conclusion: The routine assessment developed to measure the external (skin surface) sagittal spine curvatures (thoracic and lumbar) was found to be consistent with the internal curvatures (vertebral body and spinous process), and it can be employed for clinical evaluation of the spine.