Back to all papers

Radiomics features as imaging biomarkers in alveolar and craniofacial bone healing: a scoping review.

March 4, 2026pubmed logopapers

Authors

Jingzhe D,Ahmad Fauzi A,Yahya N,Nik Azis NM

Affiliations (4)

  • Faculty of Dentistry, Universiti Kebangsaan Malaysia, Kuala Lumpur, Malaysia.
  • Department of Craniofacial Diagnostic and Biosciences, Faculty of Dentistry, Universiti Kebangsaan Malaysia, Kuala Lumpur, Malaysia.
  • Centre of Diagnostic, Therapeutic and Investigative Sciences, Faculty of Health Sciences, Universiti Kebangsaan Malaysia, Kuala Lumpur, Malaysia.
  • Department of Restorative Dentistry, Faculty of Dentistry, Universiti Kebangsaan Malaysia, Kuala Lumpur, Malaysia. [email protected].

Abstract

Radiomics has emerged as a promising approach for quantifying bone regeneration by extracting high-dimensional features from routine imaging data. This scoping review aimed to map the current evidence on the application of radiomics in bone remodeling, identify methodological trends, and explore its clinical applicability in regenerative procedures. A comprehensive literature search was conducted in PubMed, Scopus, Web of Science, Elsevier, and Google Scholar up to July 2025. Eligible studies were original investigations applying radiomics features, such as texture analysis, fractal dimension, and histogram-based metrics, derived from CT or cone-beam CT (CBCT) images to assess bone healing or regenerative outcomes. Nine studies met the inclusion criteria, comprising seven retrospective and two prospective designs, with sample sizes ranging from 25 to 450 patients. All studies utilized CBCT imaging and evaluated bone healing following tooth extraction (n = 3), socket preservation (n = 2), or grafting procedures (n = 3). The most frequently extracted features were gray-level co-occurrence matrix (GLCM) parameters (n = 7), histogram of oriented gradients (HOG) (n = 3), and fractal dimensions (n = 4). Feature selection methods included least absolute shrinkage and selection operator (LASSO) regression and statistical filtering techniques. Four studies incorporated machine learning models, including 3D U-Net architectures; however, validation strategies and performance metrics were inconsistently reported. Overall, radiomics demonstrates potential for characterizing bone regeneration patterns, particularly through GLCM-based texture features. Nevertheless, validation was limited, reporting was heterogeneous, and clinical integration remains minimal. Future research should prioritize well-designed prospective studies, standardized radiomics workflows, and integration of multimodal biomarkers to facilitate reliable clinical translation.

Topics

Journal ArticleReview

Ready to Sharpen Your Edge?

Subscribe to join 11k+ peers who rely on RadAI Slice. Get the essential weekly briefing that empowers you to navigate the future of radiology.

We respect your privacy. Unsubscribe at any time.