JMIR Publications

ABSTRACT

Background:

Cervical degenerative diseases are a global public health issue, and their incidence is rising worldwide. Although an increasing number of studies on traditional machine learning (TML) and deep learning (DL) have been conducted in the detection and segmentation of cervical degenerative diseases and have reported promising task-specific results, the performance of these models has not yet been systematically analyzed.

Objective:

This systematic review and meta-analysis aimed to summarize and evaluate existing evidence on TML and DL approaches for diagnosing cervical degenerative diseases, thereby comprehensively guiding future research and clinical applications.

Methods:

This systematic review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. A comprehensive literature search was conducted on PubMed, Embase, and the Cochrane Library from January 2000 to November 2025 to identify studies evaluating TML and DL algorithms for the diagnosis of radiographic cervical degenerative diseases. The Quality Assessment of Diagnostic Accuracy Studies 2 tool was used to evaluate the risk of bias. Furthermore, the MIDAS module of Stata was employed for data synthesis and statistical analyses.

Results:

This systematic review included 18 studies, of which 14 involved a total of 12,743 patients included in the meta-analysis. Two studies were identified as having a high risk of bias, 1 as having an unclear risk, and 11 as having a completely low risk of bias. The pooled sensitivity and specificity were 0.89 (95% confidence interval [CI] = 0.86–0.91; I2 = 94.61%) and 0.88 (95% CI = 0.85–0.91; I2 = 95.87%), respectively. The positive likelihood ratio was 7.69 (95% CI = 5.89–10.03), and the negative likelihood ratio was 0.13 (95% CI = 0.10–0.16). The area under the summary receiver operating characteristic curve was 0.95 (95% CI = 0.92–0.96), indicating high diagnostic value.

Conclusions:

This systematic review and meta-analysis highlights that although artificial intelligence systems demonstrate satisfactory performance in the diagnosis of cervical degenerative diseases in experimental settings, no existing system possesses the reliability or utility required for immediate deployment in real-world clinical practice. Future endeavors should focus on the establishment of large-scale, high-quality datasets, external validation to evaluate generalizability, and the exploration of hybrid strategies merging TML with DL to bridge the significant gap between current computational models and practical clinical implementation.