Heterogeneous Modeling for Fabrication of Anatomical Structure from DICOM Image Using Additive Manufacturing

Abstract

In recent years, Additive Manufacturing (AM) unfolds its huge capability to fabricate an object in remarkable ways which either created directly from Computer Aided model and STL (Standard Tessellation Language) file or using Coordinate points and G codes. AM used in various field such as engineering, aerospace, biomedical, etc. Fused Filament Fabrication (FFF) is the most frequently used process for AM. It uses the technique of depositing a viscous material on the build platform to fabricate an object to use it for creating a mould for an industrial tool or developing models for preoperative planning for medical surgery or even Anatomical structure for demonstration purposes. FDM generally used to manufacture homogeneous object but when it comes to heterogeneous object it finds difficulty in altering desired process parameter in its software. The commercial FFF software generate toolpath whose parameters are selected by users but the choices are constrained and only limited number of variation of parameters such as raster angle, infill density, layer thickness toolpath pattern, etc. are available. Although this limitation does not affect the properties of homogenous material but fails to provide heterogeneity on a single layer. In AM certain techniques have been used in producing parts of various infill density. The optimization of infill parameter and multi contouring on a single layer gives a new method to model heterogeneous structure using FFF. A new approach has been used to obtain the density variation of human bone at its cross-section using digital image processing technique. The Developed algorithm is capable to extract useful data from an image based on density variation of bone and its variability is dependent on the value of grayscale for each pixel. Each pixel value is associated with the attenuation coefficient from the X-ray of the object which give the relative value of density at each pixel. The accumulated data correlates the heterogeneity and efficiency of the designed model with the actual bone. This single material modelling technique is more feasible and adaptable for biomedical application.