Synthetic infant head shapes with deformational plagiocephaly: concept and 3D model parameterization

• Autor(a) principal: Veloso, Fernando
• Data de Publicação: 2019
• Outros Autores: Torres, Helena R., Gomes-Fonseca, João, Oliveira, Bruno, Ruediger, Mario, Hahn, Gabriele, Pinho, António C. M., Correia-Pinto, Jorge, Morais, Pedro, Vilaça, João L.
• Tipo de documento: Artigo
• Idioma: eng
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: http://hdl.handle.net/11110/1936
• topic: deformational plagiocephaly; parametric modeling; deformational plagiocephaly; 3D head models

Deformational plagiocephaly (DP) is an asymmetric flattening of the occiput, which can be caused by a preference for a sleeping position or a restriction of the head rotation. The absence of a standardized method to quantify the severity of DP is problematic, being the severity of the condition the main factor for the recommendation of a specific treatment. To improve the quality of assessment of DP and thanks to the recent emergence of novel software solutions to process pointclouds, 3D surface cquisition of the infant’s head is performed. Although these strategies proved their added-value, they have high costs and require medical training to learn how to correctly manipulate the geometry acquisition system and the software. However, no 3D synthetic and realistic models are currently available, hampering the development of novel methods. In this sense, the future purpose of this work is to generate an artificial benchmark of 3D models of infant heads with deformational plagiocephaly and make it freely available to potentiate the development of these novel methods. Thus, a selection of the best parameters that characterize plagiocephaly is here performed. To create a parametric shape model, measurements retrieved from state-of-the-art studies and 40 clinical magnetic resonance images of infant heads (33 with normal shape and 7 with mild deformity) were extracted and evaluated. We expect to apply these parameters in a future work, on an artificial intelligence framework to generate the models.