Applied Machine Learning for Spine Surgeons: Predicting Outcome for Patients Undergoing Treatment for Lumbar Disc Herniation Using PRO Data

Abstract

Study Design: Retrospective/prospective study. Objective: Models based on preoperative factors can predict patients’ outcome at 1-year follow-up. This study measures the performance of several machine learning (ML) models and compares the results with conventional methods. Methods: Inclusion criteria were patients who had lumbar disc herniation (LDH) surgery, identified in the Danish national registry for spine surgery. Initial training of models included 16 independent variables, including demographics and presurgical patient-reported measures. Patients were grouped by reaching minimal clinically important difference or not for EuroQol, Oswestry Disability Index, Visual Analog Scale (VAS) Leg, and VAS Back and by their ability to return to work at 1 year follow-up. Data were randomly split into training, validation, and test sets by 50%/35%/15%. Deep learning, decision trees, random forest, boosted trees, and support vector machines model were trained, and for comparison, multivariate adaptive regression splines (MARS) and logistic regression models were used. Model fit was evaluated by inspecting area under the curve curves and performance during validation. Results: Seven models were arrived at. Classification errors were within ±1% to 4% SD across validation folds. ML did not yield superior performance compared with conventional models. MARS and deep learning performed consistently well. Discrepancy was greatest among VAS Leg models. Conclusions: Five predictive ML and 2 conventional models were developed, predicting improvement for LDH patients at the 1-year follow-up. We demonstrate that it is possible to build an ensemble of models with little effort as a starting point for further model optimization and selection.

AI Prediction of Neuropathic Pain after Lumbar Disc Herniation—Machine Learning Reveals Influencing Factors

“…Comparable studies on the prediction of specific therapy outcomes after spinal surgery have now been published for several years with increasing practical prediction capabilities [ 9 , 28 , 29 , 30 ]. A gradual improvement of the predictions can be observed, which we attribute to improvements in the programming possibilities of algorithms as well as an improvement in data quality [ 31 , 32 ]. Further optimization of algorithms and perfection of data sets will, from our point of view, make it possible to establish them in a way that is suitable for everyday use in the future, and will also be possible in the context of the present study.…”

Current Applications of Machine Learning in Spine: From Clinical View

“…100,101 A few models were designed for predicting outcome like, VAS, ODI, mJOA, and invasiveness score based on preoperative factors in lumbar disc herniation (LDH) or LBP patients. 14 , [102][103][104][105] Similarly, ML meaningful predicted survival outcomes of spinopelvic chondrosarcoma, ependymoma, malignant peripheral nerve sheath tumor and spinal metastases patients. [106][107][108][109][110] In spinal metastatic disease, SORG algorithms had been externally validated for survival prediction.…”

Translating Data Analytics Into Improved Spine Surgery Outcomes: A Roadmap for Biomedical Informatics Research in 2021

“… 68 , 80 These limitations, combined with the relatively simple nature of many clinical datasets, likely explain the fact that machine learning approaches have often shown modest if any advantages compared to regression in many spine clinical prediction studies. 83 - 86 Consequently, investigations using machine learning for clinical predictions should demonstrate sufficient improvements in predictive performance to justify the loss of interpretability.…”