Supervised and unsupervised learning to classify scoliosis and healthy subjects based on non-invasive rasterstereography analysis

PLoS One. 2021 Dec 23;16(12):e0261511. doi: 10.1371/journal.pone.0261511. eCollection 2021.

Abstract

The aim of our study was to classify scoliosis compared to to healthy patients using non-invasive surface acquisition via Video-raster-stereography, without prior knowledge of radiographic data. Data acquisitions were made using Rasterstereography; unsupervised learning was adopted for clustering and supervised learning was used for prediction model Support Vector Machine and Deep Network architectures were compared. A M-fold cross validation procedure was performed to evaluate the results. The accuracy and balanced accuracy of the best supervised model were close to 85%. Classification rates by class were measured using the confusion matrix, giving a low percentage of unclassified patients. Rasterstereography has turned out to be a good tool to distinguish subject with scoliosis from healthy patients limiting the exposure to unnecessary radiations.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adolescent
  • Adult
  • Female
  • Humans
  • Imaging, Three-Dimensional / methods
  • Male
  • Retrospective Studies
  • Scoliosis / classification
  • Scoliosis / diagnosis*
  • Supervised Machine Learning
  • Unsupervised Machine Learning
  • Young Adult

Grants and funding

This study was partially supported by funds awarded to MM and LP within the project Advanced Optimization-based machine learning models for analytics on clinical data (No: RM11916B7FCB6893 - 2019), which received funding from Sapienza, University of Rome. ACT Operations Research IT srl provided support in the form of a salary for TC. The specific roles of these authors are articulated in the ‘author contributions’ section. The funders had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript.