Investigation on influencing factors of ultrasonic long bone image quality and reconstruction error
Résumé
Ultrasonic imaging techniques based on the acoustic inversion theories and methodologies can be applied to reconstruct the internal structures of long bones. To verify the imaging algorithms and investigate the reconstruction discrepancy from the original images, the reflectivity models were firstly converted from the CT image and then used to generate three separate sets of simulated data using forward operator, convolution, and finite-difference methods respectively. A group of real data were also collected on the prototype bone sample. The reconstruction errors between the synthetic and predicted data were calculated, the reconstructed images were compared with the original model, and the influence factors such as aperture, noise, inversion regularization were discussed. The results demonstrate that the discrepancy is mainly located in the end and deep structures. The major reasons for the reconstruction error and information loss can be attributed to three aspects: 1) numerical truncation errors, 2) diffraction during the wave propagation, and 3) attenuation due to scattering and absorption. The compensation methods such as smaller aperture, signal processing and proper regularization constraints can be employed to improve the image quality and reduce the reconstruction errors.
Domaines
Acoustique [physics.class-ph]
Origine : Fichiers éditeurs autorisés sur une archive ouverte