Scientific Online Resource System

Scripta Scientifica Medica

Application of the deblurring technique and segmentation on medical computed tomography (CT) images for improved visualization of the cranial sutures

S. Nikolova, S. Harizanov, D. Toneva, I. Georgiev, D. Zlatareva, V. Hadjidekov, N. Lazarov

Abstract

Introduction: The investigation of the cranial suture morphology and the changes occurring during the oblit­eration is an important precondition for understanding suture maturation and its relation to ageing. The μCT-imaging of dry skulls with known age-at-death is a powerful tool in this direction, but this technology is still ex­pensive, difficult to access and entirely inapplicable on living individuals. On the other hand, the diagnostic im­aging of patients accumulates large amount of volumetric images, which could be used for scientific researches. In medical imaging, the generated data is typically artificially blurred for denoising purposes. The standard Gauss­ian blur assigns a weighted average value for the gray-scale intensity of each pixel, based on the initial intensities of the pixel and its neighbors. Unfortunately, together with the noise reduction, such a procedure decreases the contrast on image contours and significantly complicates the study of cranial sutures.

Aim: This study aims to apply theoretical and experimental scientific techniques for optimal extraction of useful data for the degree of suture obliteration, from medical tomographic images.

Materials: Volumetric images obtained from diagnostic imaging of patients with multislice CT system Toshiba Aquillion64 were used.

Methods: We developed a deblurring technique, based on the penalized total variation optimization. We also segmented the output image, using an appropriate threshold in order to separate the suture (region of interest) from the background.

Results: After processing the images, the contact between the adjacent cranial bones is clearer and the degree of suture closure could be interpreted more accurately.

Acknowledgements: This study was supported by the Bulgarian National Science Fund, grant DN01/15- 20.12.2016.





DOI: http://dx.doi.org/10.14748/ssm.v49i0.4855

Refbacks

Article Tools
Email this article (Login required)
About The Authors

S. Nikolova
IEMPAM
Bulgaria

S. Harizanov
IEMPAM
Bulgaria

D. Toneva
IEMPAM
Bulgaria

I. Georgiev
IEMPAM
Bulgaria

D. Zlatareva
IEMPAM
Bulgaria

V. Hadjidekov
IEMPAM
Bulgaria

N. Lazarov
IEMPAM
Bulgaria

Font Size


|