Spine Radiograph Analysis after Cervical Fusion: Generative Adversarial Networks as Prefixed Models Improve Postoperative Imaging Monitoring

Introduction: Postoperative radiographic parameters of interest for cervical spine surgery are currently determined manually or with computer-assisted landmark detection. Existing radiographic algorithms are sometimes hindered by vertebral radio-opaque fixation instrumentation, resulting in poor registration and landmark detection. In this work, we present a framework and model which identifies and digitally “erases” confounding instrumentation from postoperative spine images.

Methods: Vertebral end plates, intervertebral disc heights and Cobb angles, and spinous process A/P tips were landmarks of interest. Four model architectures were compared: 1) standard U-Net architecture, 2) standard registration-based model, and 3,4) both standard models prefaced with a masking U-Net and Generative Adversarial Network (GAN) that digitally erases spine instrumentation. The GAN was pre-trained on 11,694 cervical spine x-rays from the NHANES-II dataset, with digitally-added fixation hardware phantoms. These four models were evaluated on 132 annotated postoperative anterior cervical discectomy and fusion lateral cervical spine radiographs from patients at our institution between 06/2022-06/2023.

Results: Inter-rater reliability showed a normalized Euclidean Distance (ED) 1.56±0.14mm for landmark locations. The U-Net and registration models were able to locate C2-C6 vertebral parameters of interest, with an ED of 2.11±0.24mm and 1.91±0.22mm, respectively. The prefixed U-Net and registration models demonstrated ED of 1.61±0.18mm and 1.11±0.24mm, respectively. Using the tolerance benchmark of 1.56mm, precision scores were determined for each model: 88% vs. 95% for the prefixed U-Net (p = 0.0132) and 79% vs. 88% for the prefixed registration model (p = 0.0092). Subgroup analysis revealed that the prefixed models fared better (89% vs. 96%, p = 0.0004) for spinous processes vs. vertebral endplates.

Conclusion : We demonstrate that GANs trained on spinal imaging can aid neurosurgeons with postoperative cervical spine radiograph analysis, and that when used as prefixed models, they outperform state-of-the-art detection models with efficacy similar to gold standard human-detected landmarks.