Radiomics in bone pathology of the jaws.

Area of Science:

• Medical Imaging
• Radiology
• Computational Pathology

Background:

• Radiomics extracts quantitative features from medical images.
• Jawbone pathologies require accurate diagnostic and classification methods.

Purpose of the Study:

• To define radiomics features for jawbone pathologies.
• To assess their use in diagnosis, prognosis, and therapeutic response prediction.

Main Methods:

• Comprehensive literature search across eight databases and gray literature.
• Inclusion/exclusion criteria applied by two independent observers.
• 23 papers analyzed for radiomics features in jawbone pathologies.

Main Results:

• Texture, shape, and intensity features were most frequent.
• Support Vector Machine and random forest classifiers yielded best results.
• Specific features showed high performance for osteoporosis, TMJ pathology, cysts, and tumors.

Conclusions:

• Radiomics, particularly texture analysis, shows promise for jawbone pathology diagnosis.
• Features like GLCM, first-order statistics, and GLRLM are frequently used.
• Standardization is needed, but radiomics can reduce subjectivity and personalize care.