Algorithms for Adolescent Idiopathic Scoliosis Classification Based on Surface Topography Analysis

• Author / Creator

  Ghaneei, Maliheh

• While the common method for diagnosing and monitoring adolescents with idiopathic scoliosis (AIS) is X-ray radiographs from which a Cobb angle is measured, studies have shown that high radiation exposure is linked to high risk of cancer, particularly, for children and women. This thesis describes an algorithm for AIS classification based on surface topography analysis which is a radiation-free method. We present an approach which improves the user-independence level of the previously developed 3D markerless asymmetry analysis using a new asymmetry threshold without compromising its accuracy in identifying the progressive scoliosis curves. Thresholds, which have been used for separating the deformed area, were changed to automatically isolate the deformed area paired with Cobb angles. New classification trees were developed to use asymmetry parameters for classifying curve severity and progression status. In monitoring of scoliosis curves progression over a period of 12±3 months, the sensitivity of curve progression was increased from 68% to 75%, while the specificity was decreased from 74% to 59%, compared with the original method. Results demonstrate that smaller number of radiographs would be saved, however the risk of missing a curve with progression would be decreased, i.e. the proposed approach is more conservative in monitoring of scoliosis curves in clinical applications. Although using the classification tree method led to promising results, it was highly sensitive to threshold values selected in the decision trees. We demonstrate another classification algorithm, custom Neighbourhood Classifier, by which the accuracy of the curve severity and progression were increased by 17% and 58%, respectively. The new algorithm is based on the idea that curves with close asymmetry parameters are likely to belong to the same class. Regarding the contribution of each asymmetry parameters, in general, they do not play the same role in decision making, thus modification was performed to use such parameters properly.

• Subjects / Keywords

  • Adolescent Idiopathic Scoliosis (AIS)
  • Curve progression
  • Classification
  • Surface topography
  • Decision tree
  • K Nearest Neighbour
  • Monitoring
  • 3D markerless asymmetry analysis
  • Scoliosis

• Graduation date

  Spring 2018

• Type of Item

  Thesis

• Degree

  Master of Science

• DOI

  https://doi.org/10.7939/R3GH9BQ9P

• License

  This thesis is made available by the University of Alberta Libraries with permission of the copyright owner solely for non-commercial purposes. This thesis, or any portion thereof, may not otherwise be copied or reproduced without the written consent of the copyright owner, except to the extent permitted by Canadian copyright law.