(Sunday, 7/14/2019)
Room: ePoster Forums
Purpose: To obtain the bone parameters images of tissue cross-section for the quantitative detection of bone mineral density, in order to realize the early diagnosis of osteoporosis.
Methods: The modified full waveform inversion(FWI) based on the Gaussian-Newton algorithm is applied in the quantitative ultrasonic bone BMD. We adopted some 2D(dimensional) or 3D models of long bone ,tibia-fibula bone,cancellous bone and cortex bone for object models.，while the ultrasonic emission and receiver placed on the surface of the up and down of models.In the forward part,through the wave field coupling of grid points,it is used conventional grids in computational area and the staggered grids in perfectly matched layer absorbing region.In the inversion part, the parameters images of the cortical bone are obtained by multi-scale grid in frequency domain algorithm while a new joint velocity-density inversion method is used to improved the inversion effect of density.
Results: The results indicated the FWI algorithm can provide high-resolution velocity and density 3D image ,while the relative error of the numerical model is within 2%.This method has enabled velocity mapping of objects with high speed of sound contrast relative to the background medium., the speed of object bone model can be up to 4000m.s-1,and robust inversion results could also be obtained after adding Gaussian white noise.
Conclusion: This study proved that modified FWI algorithm can obtain accurate velocity parameters and density distribution image of bone quantitative ultrasonic .Numerical experiments demonstrated that our modified full-waveform inversion algorithm is effective and robust for clinical bone quantitative measurement. This method might be used in early diagnosis and treatment of bone hyperplasia successfully in future.
Keywords
BMD, Ultrasonics, Quantitative Imaging
Taxonomy
IM/TH- Image Analysis (Single modality or Multi-modality): Quantitative imaging
Contact Email
