Fuzzy-logic-assisted surgical planning in adolescent idiopathic scoliosis

J Spinal Disord Tech. 2009 Jun;22(4):263-9. doi: 10.1097/BSD.0b013e3181761950.

Abstract

Summary of background data: Selection of appropriate curve fusion levels for surgery in adolescent idiopathic scoliosis (AIS) is a complex and difficult task and, despite numerous publications, still remains a highly controversial topic.

Objective: To evaluate a fuzzy-logic-based surgical planning tool by comparing the results suggested by the software with the average outcome recommended by a panel of 5 expert spinal deformity surgeons. It is hypothesized that, given the same information, the fuzzy-logic tool will perform as favorably as the surgeons.

Study design: Proof-of-concept study evaluating the use of a fuzzy-logic-assisted surgical planning tool in AIS to select the appropriate spinal curve to be instrumented.

Methods: A cohort of 30 AIS surgical cases with a main thoracic curve was used. Each case included standard measurements recorded from preoperative standing postero-anterior and lateral, supine side bending, and 1-year postoperative standing radiographs. Five experienced spinal deformity surgeons evaluated each case independently and gave their preferred levels of instrumentation and fusion. The cases were then presented to the fuzzy-logic tool to determine whether the high thoracic and/or lumbar curves were to be instrumented. For each case, a percentage value was obtained indicating inclusion/exclusion of the respective curves in the surgical instrumentation procedure. Kappa statistics was used to compare the model output and the average decision of the surgeons.

Results: Kappa values of 0.71 and 0.64 were obtained, respectively, for the proximal thoracic and lumbar curves models, thus suggesting a good agreement of the fusion recommendations made by the fuzzy-logic tool and the surgeons.

Conclusions: Given the same information, the fuzzy-logic-assisted recommendation of the curve to be instrumented compared favorably with the collective decision of the surgeons. The findings thus suggest that a fuzzy-logic approach is helpful in assisting surgeons with the preoperative selection of curve instrumentation and fusion levels in AIS.

Publication types

  • Clinical Trial
  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adolescent
  • Female
  • Fuzzy Logic*
  • Humans
  • Male
  • Scoliosis / surgery*
  • Spinal Fusion / instrumentation*
  • Spinal Fusion / methods*
  • Surgery, Computer-Assisted / methods*
  • Treatment Outcome