Aims. Clinical decision support systems (CDSS) can support clinicians in selecting appropriate treatments for patients. The objective of this study was to examine if triaging patients with LBP to the most optimal treatment can be improved by using a data-driven approach with the help of machine learning as base of such a CDSS. Methods. A clinical database of the Groningen Spine Center containing patient-reported data from 1546 patients with LBP was used. From this dataset, a training dataset with 354 features was labeled on eight different treatments actually received by these patients. With this dataset, models were trained. A test dataset with 50 cases judged on treatments by 4 experts in LBP triage was used to test these models with data not used to train the models. Prediction accuracy and average area under curve (AUC) were used as performance measures for the models. Results. The AUC values indicated small to medium learning effects showing that machine learning on patient-reported data, to model decision-making processes on treatments for LBP, may be possible. One of the best performing models was the Bayesian Network (BN) model; e.g. predicted surgery with accuracy 0.78 (95% C.I. 0.68– 0.87) and AUC 0.70. Conclusion. Benefits to using BNs compared to other supervised machine learning techniques are that it is easy to exploit expert knowledge in BN models, meaning that advices generated by the model can be explained. The next step is to improve the BN accuracy so that it can actually be used in a CDSS. No conflicts of interest. Sources of funding: This work is partly funded by a grant from the Netherlands Organization for Health Research and Development (ZonMw), grant 10-10400-98-009

Aims. To determine the major risk factors for unplanned reoperations (UROs) following corrective surgery for adult spinal deformity (ASD) and their interactions, using machine learning-based prediction algorithms and game theory. Methods. Patients who underwent surgery for ASD, with a minimum of two-year follow-up, were retrospectively reviewed. In total, 210 patients were included and randomly allocated into training (70% of the sample size) and test (the remaining 30%) sets to develop the machine learning algorithm. Risk factors were included in the analysis, along with clinical characteristics and parameters acquired through diagnostic radiology. Results. Overall, 152 patients without and 58 with a history of surgical revision following surgery for ASD were observed; the mean age was 68.9 years (SD 8.7) and 66.9 years (SD 6.6), respectively. On implementing a random forest model, the classification of URO events resulted in a balanced accuracy of 86.8%. Among machine learning-extracted risk factors, URO, proximal junction failure (PJF), and postoperative distance from the posterosuperior corner of C7 and the vertical axis from the centroid of C2 (SVA) were significant upon Kaplan-Meier survival analysis. Conclusion. The major risk factors for URO following surgery for ASD, i.e. postoperative SVA and PJF, and their interactions were identified using a machine learning algorithm and game theory. Clinical benefits will depend on patient risk profiles. Cite this article: Bone Joint Res 2023;12(4):245–255

Background. Magnetic resonance imaging (MRI) algorithm identifies end stage severely degenerated disc as ‘black’, and a moderately degenerate to non-degenerated disc as ‘white’. MRI is based on signal intensity changes that identifies loss of proteoglycans, water, and general radial bulging but lacks association with microscopic features such as fissure, endplate damage, persistent inflammatory catabolism that facilitates proteoglycan loss leading to ultimate collapse of annulus with neo-innervation and vascularization, as an indicator of pain. Thus, we propose a novel machine learning based imaging tool that combines quantifiable microscopic histopathological features with macroscopic signal intensities changes for hybrid assessment of disc degeneration. Methods. 100-disc tissue were collected from patients undergoing surgeries and cadaveric controls, age range of 35–75 years. MRI Pfirrmann grades were collected in each case, and each disc specimen were processed to identify the 1) region of interest 2) analytical imaging vector 3) data assimilation, grading and scoring pattern 4) identification of machine learning algorithm 5) predictive learning parameters to form an interface between hardware and software operating system. Results. Kernel algorithm defines non-linear data in xy histogram. X,Y values are scored histological spatial variables that signifies loss of proteoglycans, blood vessels ingrowth, and occurrence of tears or fissures in the inner and outer annulus regions mapped with the dampening and graded series of signal intensity changes. Conclusion. To our knowledge this study is the first to propose a machine learning method between microscopic spatial tissue changes and macroscopic signal intensity grades in the intervertebral disc. No conflict of interest declared.  . Sources of Funding. ICMR/5/4-5/3/42/Neuro/2022-NCD-1, Dr TMA PAI SMU/ 131/ REG/ TMA PURK/ 164/2020. A part of the above study was presented as an oral paper at the International Society for the Study of Lumbar Spine (ISSLS) meeting held on 1–5. th. May 2023, Melbourne, Australia

Purposes and Background. Musculoskeletal disorders including as back and neck pain are leading causes of work disability. Effective interventions exist (i.e. functional restoration, multidisciplinary biopsychosocial rehabilitation, workplace-based interventions, etc.), but it is difficult to select the optimal intervention for specific patients. The Work Assessment Triage Tool (WATT) is a clinical decision support tool developed using machine learning to help select interventions. The WATT algorithm categorizes patients based on individual, occupational, and clinical characteristics according to likelihood of successful return-to-work following rehabilitation. Internal validation showed acceptable classification accuracy, but WATT has not been tested beyond the original development sample. Our purpose was to externally validate the WATT. Methods and Results. A population-based cohort design was used, with administrative and clinical data extracted from a Canadian provincial compensation database. Data were available on workers being considered for rehabilitation between January 2013 and December 2016. Data was obtained on patient characteristics (ie. age, sex, education level), clinical factors (ie. diagnosis, part of body affected, pain and disability ratings), occupational factors (ie. occupation, employment status, modified work availability), type of rehabilitation program undertaken, and return-to-work outcomes (receipt of wage replacement benefits 30 days after assessment). Analysis included classification accuracy statistics of WATT recommendations for selecting interventions that lead to successful RTW outcomes. The sample included 5296 workers of which 33% had spinal conditions. Sensitivity of the WATT was 0.35 while specificity was 0.83. Overall accuracy was 73%. Conclusion. Accuracy of the WATT for selecting successful rehabilitation programs was modest. Algorithm revision and further validation is needed. No conflicts of interest. Sources of funding: Funding was provided by the Workers' Compensation Board of Alberta

Aims

This systematic review aims to identify 3D predictors derived from biplanar reconstruction, and to describe current methods for improving curve prediction in patients with mild adolescent idiopathic scoliosis.