Isotope computed tomography using cone-beam geometry: a comparison of two reconstruction algorithms

Area of Science:

• Medical Imaging
• Biomedical Engineering
• Radiology

Background:

• Accurate assessment of bone mineral distribution is crucial for diagnosing and managing bone diseases.
• Existing methods may lack precision in evaluating medullary bone mineral content.

Purpose of the Study:

• To develop and evaluate a specialized CT scanner for determining the 3D bone mineral distribution in the medullary cavities of the radius, ulna, and femur.
• To compare the efficacy of Maximum Entropy (ME) and Convolution Back Projection (CBP) reconstruction methods for this application.

Main Methods:

• Construction of a dedicated CT scanner with a 153Gd x-ray source and a multiwire proportional counter (MWPC).
• Acquisition of 2D transmission projections over 360 degrees with the limb positioned axially.
• Reconstruction of bone mineral distribution using ME and CBP algorithms, evaluated with a simulated forearm phantom.

Main Results:

• Both ME and CBP methods successfully reconstructed bone mineral distribution.
• ME achieved a systematic error <3.5% and precision <2.5% for mean medullary densities of ulna and radius.
• CBP yielded a systematic error <11%, reducible to <2% with more views, and precision <3.5%.

Conclusions:

• The developed CT scanner is effective for quantifying 3D bone mineral distribution in medullary cavities.
• The Maximum Entropy method offers superior accuracy and precision compared to Convolution Back Projection for this specific application.
• This technology has potential for improved diagnosis and monitoring of bone conditions affecting medullary bone.