Automatic identification and validation of planar collagen organization in the aorta wall with application to abdominal aortic aneurysm
2013 (English)In: Microscopy and Microanalysis, ISSN 1431-9276, E-ISSN 1435-8115, Vol. 19, no 6, 1395-1404 p.Article in journal (Refereed) Published
Arterial physiology relies on a delicate three-dimensional (3D) organization of cells and extracellular matrix, which is remarkably altered by vascular diseases like abdominal aortic aneurysms (AAA). The ability to explore the micro-histology of the aorta wall is important in the study of vascular pathologies and in the development of vascular constitutive models, i.e., mathematical descriptions of biomechanical properties of the wall. The present study reports and validates a fast image processing sequence capable of quantifying collagen fiber organization from histological stains. Powering and re-normalizing the histogram of the classical fast Fourier transformation (FFT) is a key step in the proposed analysis sequence. This modification introduces a powering parameter w, which was calibrated to best fit the reference data obtained using classical FFT and polarized light microscopy (PLM) of stained histological slices of AAA wall samples. The values of w = 3 and 7 give the best correlation (Pearson's correlation coefficient larger than 0.7, R 2 about 0.7) with the classical FFT approach and PLM measurements. A fast and operator independent method to identify collagen organization in the arterial wall was developed and validated. This overcomes severe limitations of currently applied methods like PLM to identify collagen organization in the arterial wall.
Place, publisher, year, edition, pages
2013. Vol. 19, no 6, 1395-1404 p.
collagen dispersion, collagen organization, fast Fourier transform, polarized light microscopy
Medical and Health Sciences Engineering and Technology
IdentifiersURN: urn:nbn:se:kth:diva-139888DOI: 10.1017/S1431927613013251ISI: 000330344800002ScopusID: 2-s2.0-84888192428OAI: oai:DiVA.org:kth-139888DiVA: diva2:689901
FunderSwedish Research Council, 2010-4446
QC 201401222014-01-222014-01-152014-02-27Bibliographically approved