Automatic Quantification of Spinal Curvature in Scoliotic Radiograph using Image Processing

Authors:
Anitha H;G. K. Prabhu
Affiliations:
Department of Electronics and Communication Engineering, Manipal Institute of Technology, Manipal University, Manipal, India 576 104;Department of Bio-Medical Engineering, Manipal Institute of Technology, Manipal University, Manipal, India 576 104
Venue:
Journal of Medical Systems
Year:
2012

Citing 2
Cited 0

Scale-Space and Edge Detection Using Anisotropic Diffusion

IEEE Transactions on Pattern Analysis and Machine Intelligence
Snakes, shapes, and gradient vector flow

IEEE Transactions on Image Processing

Quantified Score

Hi-index	0.00

Visualization

Abstract

Choosing the most suitable treatment for the scoliosis relies heavily on accurate and reproducible spinal curvature measurement from radiographs. Our objective is to reduce the variability in spinal curvature measurement by reducing the user intervention and bias. In order to determine the reliability of the spinal curvature measurement as it is in the clinical measurement of scoliosis a methodological survey has been carried out that concludes with inter and intra observer error variation. The proposed method list out horizontal inclination of all the vertebrae's in terms of slopes using active contour models and morphological operators. This facilitates the radiologist to decide end vertebrae and hence inter/intra observer variation is completely eliminated. Tables 1 and 2 shows the observer error variation between manual and proposed methods in terms of mean and standard deviation.