Principal component and factor analysis to study variations in the aging lumbar spine

IEEE J Biomed Health Inform. 2015 Mar;19(2):745-51. doi: 10.1109/JBHI.2014.2328433. Epub 2014 Jun 2.

Abstract

Human spine is a multifunctional structure of human body consisting of bones, joints, ligaments, and muscles which all undergo a process of change with the age. A sudden change in these features either naturally or through injury can lead to some serious medical conditions which puts huge burden on health services and economy. While aging is inevitable, the effect of aging on different areas of spine is of clinical significance. This paper reports the growth and degenerative pattern of human spine using principal component analysis. Some noticeable lumbar spine features such as vertebral heights, disc heights, disc signal intensities, paraspinal muscles, subcutaneous fats, psoas muscles, and cerebrospinal fluid were used to study the variations seen on lumbar spine with the natural aging. These features were extracted from lumbar spine magnetic resonance images of 61 subjects with age ranging from 2 to 93 years. Principal component analysis is used to transform complex and multivariate feature space to a smaller meaningful representation. PCA transformation provided 2-D visualization and knowledge of variations among spinal features. Further useful information about correlation among the spinal features is acquired through factor analysis. The knowledge of age related changes in spinal features are important in understanding different spine related problems.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adolescent
  • Adult
  • Aged
  • Aged, 80 and over
  • Aging / pathology*
  • Child
  • Child, Preschool
  • Factor Analysis, Statistical
  • Female
  • Humans
  • Image Processing, Computer-Assisted / methods*
  • Lumbar Vertebrae / pathology*
  • Lumbosacral Region / physiology
  • Magnetic Resonance Imaging
  • Male
  • Middle Aged
  • Muscle, Skeletal / physiology
  • Principal Component Analysis
  • Young Adult