Diffusion MRI streamline tractography offers a unique approach to probe into such mechanisms underlying structural brain network. In this study, we propose modelling network strength based on both fibre length and topological profile of brain network. The proposed model leads to more robust fitting outcomes. Meanwhile, our results show that network thresholding could dramatically alter the relationship between the connection strength and physical length; this would inevitably compromise further network analyses. Further results demonstrate that within- and between-hemisphere connections bear distinct patterns in terms of the relationship between network strength and topological correlation matrix.