Machine learning can help predict early-onset adjacent segment degeneration following ACDF

6992
Samuel Rudisill (left) and Dino Samartzis (right)

Machine learning (ML) can help in the discovery of early-onset adjacent segment degeneration (EO-ASD) as well as predict its development following anterior cervical discectomy and fusion (ACDF). As such, the technology can be used to support clinical decision-making and precision-based spine care.

This is according to recent research, published in the European Spine Journal by Samuel Rudisill, Dino Samartzis and their multidisciplinary team at Rush University Medical Center (Chicago, USA), in conjunction with colleagues from Ulm University (Ulm, Germany) and Virginia Mason Medical Center (Seattle, USA).

The study aimed to develop and validate a machine learning model to predict EO-ASD following ACDF.

A retrospective review of prospectively collected data of patients undergoing ACDF at a quaternary referral medical centre was performed. Patients over 18 years of age with more than six months of follow-up and complete pre- and postoperative plain radiographs and magnetic resonance imaging (MRI) were included in the study.

A machine learning-based algorithm was developed to predict EO-ASD based on preoperative demographic, clinical, and imaging parameters, and model performance was evaluated according to discrimination and overall performance.

A total of 366 ACDF patients were included (50.8% male; mean age 51.4 ± 11.1 years). Over 18.7 ± 20.9 months of follow-up, 97 (26.5%) patients developed EO-ASD.

The researchers found that the machine learning model demonstrated good discrimination and overall performance according to precision (EO-ASD: 0.70, non-ASD: 0.88), recall (EO-ASD: 0.73, non-ASD: 0.87), accuracy (0.82), F1-score (0.79), Brier score (0.203), and area under the ROC curve (AUC) (0.794).

In addition, specific patient phenotypes were identified as the most important predictive features of EO-ASD following a cervical fusion.

Speaking to Spinal News International, Samartzis said: “The field, analytically, has progressed substantially. Bringing together clinicians, data scientists and computer engineer experts, we’ve been able to provide new insights on spine degeneration that has clinical relevance. ‘Intelligence-based spine care’ is here to stay but it is up to the collective community to aim for quality to lend itself to more robust personalized spine care approaches. In the setting of this study, we hope to further continue to validate our model, raise awareness and discussion in an effort to refine outcome prediction and ultimately improve patient care.”


LEAVE A REPLY

Please enter your comment!
Please enter your name here