Anterior cervical discectomy and fusion (ACDF) is one of the most commonly performed surgical procedures for the treatment of cervical degenerative disease, myelopathy, radiculopathy, and segmental instability. Although clinical outcomes are generally favorable, pseudarthrosis remains a relevant complication, with a reported incidence ranging from 5% to 20%. In a field with no yet clear main directions, this narrative review aims at giving the reader a broad picture and a wide analysis of the recent advances in cervical spinal fusion, with particular focus on biomaterials, intervertebral cage technologies, cervical spine biomechanics and imaging methods used for fusion assessment. The literature regarding quantitative imaging parameters and emerging applications of artificial intelligence (AI) is also reviewed. Current bone grafts include autologous grafts, allografts, xenografts and polymeric grafts, while the materials for the intervertebral cages comprehend titanium, polyetheretherketone and silicon nitride, with reported fusion rates distributed in a very large range. Computed tomography (CT) remains the standard imaging modality to assess whether fusion has occurred, due to its high spatial resolution. However, the lack of shared diagnostic criteria and the significant interobserver variability continue to limit its reliability. Quantitative parameters, such as Hounsfield Unit measurements and MRI-derived bone quality scores, may contribute to a more objective evaluation, although current evidence remains heterogeneous. In parallel, AI-based imaging analysis is showing promising results for quantitative assessment and longitudinal monitoring of bone fusion; however, large prospective clinical studies are still needed to confirm its clinical applicability. In conclusion, despite advances in surgical technologies and biomaterials, radiological assessment of cervical fusion still lacks universally accepted diagnostic standards. Future AI applications may improve diagnostic accuracy and reproducibility, promoting a more standardized approach in clinical practice.
Evangelisti et al. (Mon,) studied this question.