Improved algorithms for automatic bone histomorphometry on a numerized image analysis system

Abstract : Quantitative histological methods have proved to be the most effective methods in bone disease research. Faster and more accurate techniques are currently needed. We have developed a simple digitized image analysis system which allows accurate measurements of trabecular bone mass. The algorithm is based on the 'four-connected sets' mathematical theory. Given a numerized image displayed by a CCD camera, the algorithm recognizes all possible four-connected sets and provides area measurements. The first procedure automatically eliminates small, irrelevant profiles (wrinkles, cell nuclei, etc.) while larger profiles are erased interactively. The second procedure similarly erases the artefactual defects within the trabecule (artefactual cracks or empty osteocytic lacunae). The method was shown to be very accurate and time-saving.
Type de document :
Article dans une revue
Journal of Microscopy, Wiley, 1989, Vol 150 (Issue 2), p. 151-160. 〈10.1111/j.1365-2818.1988.tb04605.x〉
Liste complète des métadonnées

https://hal-emse.ccsd.cnrs.fr/emse-00947807
Contributeur : Florent Breuil <>
Soumis le : lundi 17 février 2014 - 14:19:08
Dernière modification le : mardi 22 mars 2016 - 01:15:58

Identifiants

Citation

Michel Beigbeder, D. Chappard, C. Alexandre, L. Vico, S. Palle, et al.. Improved algorithms for automatic bone histomorphometry on a numerized image analysis system. Journal of Microscopy, Wiley, 1989, Vol 150 (Issue 2), p. 151-160. 〈10.1111/j.1365-2818.1988.tb04605.x〉. 〈emse-00947807〉

Partager

Métriques

Consultations de la notice

47