Numerical characterisation of biomedical titanium surface texture using novel feature parameters

In biomedicine, titanium materials with specific surface textures (for example, those produced using consecutive polishing, sandblasting and acid etching) are widely used to facilitate osseointegration of human/animal tissue with implant surface structures. The surface texture has a critical role on the material's functionality in terms of cellular adhesion and proliferation. However, conventional surface topography characterisation parameters pertinent to cellular attachment such as Sds (density of summits of a surface) and Ssc (arithmetic mean summit curvature of a surface) are liable to be influenced by low amplitude high spatial frequency components or measurement noise. In this research, a novel feature characterisation, based on pattern recognition techniques is implemented on specially processed titanium surfaces. The mean dimensions and densities of the micro-scale features of the surfaces are extracted. The statistical analysis results demonstrate the efficiency and stability of the feature analysis compared to the use of conventional surface texture parameters. Additionally, potentially efficient parameters for characterisation of biomedical surfaces are indicated through the use of a one-way analysis of variance. As a case study, from the point of view of surface metrology, limited experimental data is presented; the intention of the authors is to give a guide to innovative use of the novel surface characterisation techniques. A large amount of biomedical experiments would be needed in the future to fully validate the correlation between the surface texture parameters and its biomedical functions but the present work provides a useful start point for a larger study.

► Feature characterisation is firstly used in bio-medical material measurements.
► Stabilities of feature parameters and field parameters are demonstrated.
► Feature size-related parameters of the bio-material are extracted.
► Potential functionally correlated parameters are selected.