Automated Clustering Technique (ACT) for Early Onset Scoliosis

Girish Viraraghavan ¹, Patrick J. Cahill², Michael G. Vitale³, Brendan A. Williams², Pediatric Spine Study Group⁴, Sriram Balasubramanian¹

¹School of Biomedical Engineering, Science and Health Systems, Drexel University, Philadelphia, PA, ²The Children's Hospital of Philadelphia, Philadelphia, PA ³Division of Pediatric Orthopaedic Surgery, Department of Orthopaedic Surgery, Columbia University, Morgan Stanley Children's Hospital of New York-Presbyterian, Columbia University Medical Center, New York, NY, ⁴Valley Forge, PA, USA

This is an application created to cluster Early Onset Scoliosis (EOS) patients based on the ACT-EOS system.

1170 EOS patients retrospectively selected from the Pediatric Spine Study Group (PSSG) database were used for the clustering analyses. The clinical indices used for the automated Fuzzy C-means clustering analysis were age, major curve (Cobb) angle, kyphosis, number of levels involved in the major curve (Cobb) angle, number of levels involved in kyphosis and deformity index (defined as the ratio of major curve (Cobb) angle and kyphosis).

Please refer to Table I for Etiology,range of values for Cobb angle and kyphsis for each etiology

Etiology:
Cobb angle (degrees):
Kyphosis (degrees):

Place the patient in a cluster

Results:

Table A: Etiology,range of values for Cobb angle and kyphsis for each etiology