Prediction of post-operative clinical parameters in posterior scoliosis surgery through an adaptive neuro-fuzzy interface system

Background and Objective: Post-operative clinical indices should be estimated accurately in scoliosiscorrection surgeries. It has been analyzed in various studies such as experimental (in vitro orin vivo) trials through different modeling methods (finite element or multi-body analysis). Thesecostly and time-consuming methods can only be applied to a large population of scoliotic patients.An adaptive neuro-fuzzy interface system (ANFIS) is used in this study to estimate the post-operativecobb and thoracic kyphosis angles in adolescent idiopathic scoliosis patients undergoingposterior scoliosis correction surgeries.Methods: Four groups of ۵۵ patients with distinct pre-operative clinical indices (thoracic cobband pelvic incidence) were use۱d the ANFIS inputs, whereas posto-perative thoracic cobb andkyphosis angles were utilized as the outputs. For robustness evaluation, the predicted values ofpost-operative angles were compared with measurements by calculating the root mean squareerrors and clinical correction deviation indices (the relative deviation of post-operative predictedangles from the real angles).Results: The least root mean square errors (۳.۰º and ۶.۳° for the main thoracic cobb and thoracickyphosis estimations, respectively) were recorded in the group with the main thoracic cobb,pelvic incidence, thoracic kyphosis, and T۱ spinopelvic inclination used as inputs. The clinicalcorrection deviation indices were calculated ۰.۰۰۸۶ and ۰.۰۶۴۱ for cobb angles in two cases and۰.۰۵۳۴ and ۰.۲۸۷۹ for thoracic kyphosis in two other cases.Conclusion: Greater differences between pre-operative and post-operative cobb angles comparedwith those of thoracic kyphosis decreased the root-mean-square errors and clinical deviation indicesbut improved accuracy.

کلیدواژه ها:

Cobb angle ، Pelvic incidence ، Scoliosis correction ، Spine ، Thoracic kyphosis