Using texture analysis of head CT images to differentiate osteoporosis from normal bone density

Area of Science:

• Radiology
• Medical Imaging
• Osteoporosis Research

Background:

• Osteoporosis diagnosis typically relies on dual-energy X-ray absorptiometry (DXA).
• Noncontrast head CT scans are common in clinical practice.
• There is a need for alternative osteoporosis detection methods.

Purpose of the Study:

• To evaluate texture analysis for osteoporosis detection on head CTs.
• To identify optimal regions for bone sampling within craniofacial bones.

Main Methods:

• Retrospective study of 58 patients (29 osteoporosis, 29 normal bone density).
• Manual segmentation of clivus, sphenoid triangles, and mandibular condyles on head CTs.
• Extraction and statistical analysis of 41 texture features.

Main Results:

• Significant differences in texture features between osteoporosis and normal bone density groups were found.
• Multiple texture features in the clivus, sphenoid triangles, and mandibular condyles showed significant variations.
• The number of significant features varied by anatomical region, with the left condyle showing 31 features.

Conclusions:

• Texture analysis of specific craniofacial bone regions can differentiate between normal bone density and osteoporosis.
• This technique shows potential for identifying osteoporosis on head CT scans.
• It may aid in screening patients who have not undergone DXA.