Automatic method for determining the number of lumbar and thoracic vertebrae in rabbits using Computer Tomography images

There are several studies dealing with the phenotypic variance of the vertebral number in the spinal column of rabbits. According to the literature the number of thoracic and lumbar vertebrae varies between 11-13 and 6-8, respectively. The length of the m. longissimus dorsi (MLD) - a valuable meat p...

Full description

Saved in:
Bibliographic Details
Published inActa Agraria Kaposváriensis Vol. 25; no. 2; pp. 41 - 51
Main Authors Csóka, Ádám, Petneházy, Örs, Fajtai, Dániel, Sándor, Máté, Orsi-Gibicsár, Szilvia, Donkó, Tamás
Format Journal Article
LanguageEnglish
Published 15.12.2021
Online AccessGet full text

Cover

Loading…
More Information
Summary:There are several studies dealing with the phenotypic variance of the vertebral number in the spinal column of rabbits. According to the literature the number of thoracic and lumbar vertebrae varies between 11-13 and 6-8, respectively. The length of the m. longissimus dorsi (MLD) - a valuable meat part of rabbits - is determined by the length of the vertebral column therefore the number of vertebrae may have economic importance in breeding. The aim of this study was to create an automatic counter using computed tomography (CT) images. In the first step, a skeleton binary mask was created using the radiodensity range between 120 and 3071 HU, then the lumbar and thoracic regions were processed by two different methods. The lumbar part was evaluated based on the frequency of the bone voxels along the axial plane. The number of thoracic vertebrae was determined from the number of ribs. The left and right ribs were processed separately. The developed method was tested on CT examination of 40 Hycole rabbits compared to manual evaluation. The results of the automatic algorithm had few errors: in one case in the lumbar region (2.5%) and in 3 cases in the thoracic region (5%). The automated evaluation process takes a few seconds per individual and then the program visualizes the results on a graph. The incorrectly evaluated rabbits are recognizable on graphs and they can be easily corrected with a minimal time investment.
ISSN:1418-1789
DOI:10.31914/aak.2626