Predicting radiographic outcomes of vertebral body tethering in adolescent idiopathic scoliosis patients using machine learning.

Anterior Vertebral Body Tethering (AVBT) is a growing alternative treatment for adolescent idiopathic scoliosis (AIS), offering an option besides spinal fusion. While AVBT aims to correct spinal deformity through growth correction, its outcomes have been mixed. To improve surgical outcomes, this study aimed to develop a machine learning-based tool to predict short- and midterm spinal curve correction in AIS patients who underwent AVBT surgery, using the most predictive clinical, radiographic, and surgical parameters. After institutional review board approval and based on inclusion criteria, 91 AIS patients who underwent AVBT surgery were selected from the Shriners Hospitals for Children, Philadelphia. For all patients, longitudinal standing (PA or AP, and lateral) and side bending spinal Radiographs were retrospectively obtained at six visits: preop and first standing, one year, two years, five years postop, and at the most recent follow-up. Demographic, radiographic, and surgical features associated with curve correction were collected. The sequential backward feature selection method was used to eliminate correlated features and to provide a rank-ordered list of the most predictive features of the AVBT correction. A Gradient Boosting Regressor (GBR) model was trained and tested using the selected features to predict the final correction of the curve in AIS patients. Eleven most predictive features were identified. The GBR model predicted the final Cobb angle with an average error of 6.3 ± 5.6 degrees. The model also provided a prediction interval, where 84% of the actual values were within the 90% prediction interval. A list of the most predictive features for AVBT curve correction was provided. The GBR model, trained on these features, predicted the final curve magnitude with a clinically acceptable margin of error. This model can be used as a clinical tool to plan AVBT surgical parameters and improve outcomes.

Predicting curve progression for adolescent idiopathic scoliosis using random forest model.

BACKGROUND: Adolescent Idiopathic Scoliosis (AIS) is a three-dimensional (3D) spinal deformity characterized by coronal curvature and rotational deformity. Predicting curve progression is important for the selection and timing of treatment. Although there is a consensus in the literature regarding prognostic factors associated with curve progression, the order of importance, as well as the combination of factors that are most predictive of curve progression is unknown. OBJECTIVES: (1) create an ordered list of prognostic factors that most contribute to curve progression, and (2) develop and validate a Machine Learning (ML) model to predict the final major Cobb angle in AIS patients. METHODS: 193 AIS patients were selected for the current study. Preoperative PA, lateral and lateral bending radiographs were retrospectively obtained from the Shriners Hospitals for Children. Demographic and radiographic features, previously reported to be associated with curve progression, were collected. Sequential Backward Floating Selection (SBFS) was used to select a subset of the most predictive features. Based on the performance of several machine learning methods, a Random Forest (RF) regressor model was used to provide the importance rank of prognostic features and to predict the final major Cobb angle. RESULTS: The seven most predictive prognostic features in the order of importance were initial major Cobb angle, flexibility, initial lumbar lordosis angle, initial thoracic kyphosis angle, age at last visit, number of levels involved, and Risser "+" stage at the first visit. The RF model predicted the final major Cobb angle with a Mean Absolute Error (MAE) of 4.64 degrees. CONCLUSION: A RF model was developed and validated to identify the most important prognostic features for curve progression and predict the final major Cobb angle. It is possible to predict the final major Cobb angle value within 5 degrees error from 2D radiographic features. Such methods could be directly applied to guide intervention timing and optimization for AIS treatment.

Part 2. Review and meta-analysis of studies on modulation of longitudinal bone growth and growth plate activity: A micro-scale perspective.

Macro-scale changes in longitudinal bone growth resulting from mechanical loading were shown in Part 1 of this review to depend on load magnitude, anatomical location, and species. While no significant effect on longitudinal growth was observed by varying frequency and amplitude of cyclic loading, such variations, in addition to loading duration and species, were shown to affect the morphology, viability, and gene and protein expression within the growth plate. Intermittent compression regimens were shown to preserve or increase growth plate height while stimulating increased chondrocyte presence in the hypertrophic zone relative to persistent and static loading regimens. Gene and protein expressions related to matrix synthesis and degradation, as well as regulation of chondrocyte apoptosis were shown to exhibit magnitude-, frequency-, and duration-dependent responses to loading regimen. Chondrocyte viability was shown to be largely preserved within physiological bounds of magnitude, frequency, amplitude, and duration. Persistent static loading was shown to be associated with overall growth plate height in tension only, reducing it in compression, while affecting growth plate zone heights differently across species and encouraging mineralization relative to intermittent cyclic loading. Lateral loading of the growth plate, as well as microfluidic approaches are relatively understudied, and age, anatomical location, and species effects within these approaches are undefined. Understanding the micro-scale effects of varied loading regimes can assist in the development of growth modulation methods and device designs optimized for growth plate viability preservation or mineralization stimulation based on patient age and anatomical location.

Part 1. Review and meta-analysis of studies on modulation of longitudinal bone growth and growth plate activity: A macro-scale perspective.

Growth modulation is an emerging method for treatment of angular skeletal deformities such as adolescent idiopathic scoliosis (AIS). The Hueter-Volkmann law, by which growth is stimulated in tension and inhibited in compression, is widely understood, and applied in current growth-modulating interventions such as anterior vertebral body tethering (AVBT) for AIS. However, without quantification of the growth rate effects of tension or compression, the possibility of under- or over- correction exists. A definitive mechanical growth modulation relationship relating to treatment of such skeletal deformities is yet to exist, and the mechanisms by which growth rate is regulated and altered are not fully defined. Review of current literature demonstrates that longitudinal (i.e., lengthwise) growth rate in multiple animal models depend on load magnitude, anatomical location, and species. Additionally, alterations in growth plate morphology and viability vary by loading parameters such as magnitude, frequency, and whether the load was applied persistently or intermittently. The aggregate findings of the reviewed studies will assist in work towards increasingly precise and clinically successful growth modulation methods. Part 1 of this review focuses on the effects of mechanical loading, species, age, and anatomical location on the macro-scale alterations in longitudinal bone growth, as well as factors that affect growth plate material properties. Part 2 considers the effects on micro-scale alterations in growth plate morphology such as zone heights and proportions, chondrocyte viability, and related gene and protein expression.

Thoracic vertebral morphology in normal and scoliosis deformity in skeletally immature rabbits: A Longitudinal study.

OBJECTIVE: To measure age-related changes in thoracic vertebral body heights (VBH) in skeletally immature normative and scoliotic rabbits to assess how VBH change during growth. To examine the potential link between the moment-arm of the rib tether and vertebral wedging as well as the sum of the curvature angles at the apical level (T7). To assess the correlation between the magnitude of initial spine curve and final spine curve in the scoliotic group. METHODS: Eight healthy, skeletally immature normative New Zealand rabbits and ten skeletally immature scoliotic rabbits which underwent unilateral rib tethering were included retrospectively. Each rabbit was scanned at two to four time points (at 7, 11, 14 and 28 weeks). Three dimensional bone models of thoracic vertebrae (T1-T12) were digitally segmented and reconstructed. VBH were calculated using surface landmark points from each thoracic vertebra. Apical level (T7) ± 2 levels in scoliotic rabbits were compared to their corresponding levels and time points in the normative group. The moment-arms between the centroids of 2D projections of T3-T9 vertebral bodies and the line which connects the centroids of the end levels were calculated. RESULTS: Bilateral left-right (L-R) symmetry and anterior-posterior (A-P) asymmetry were observed in normative VBH. Bilateral concave-convex (CC-CX) asymmetry and (A-P) asymmetry were observed in scoliotic VBH. No significant differences in growth rates were found between the normative and scoliotic groups. Vertebral wedging as well as curvature magnitude were positively correlated with the moment-arms. CONCLUSION: Unilateral rib tether applies compressive forces on both concave and convex sides, whereas compressive forces are lower on the latter. Knowing the amount of vertebral wedging or curve magnitude would enable us to predict the applied force (moment-arms), which is important for planning a corrective surgery.