Tomographic image reconstruction from limited projections using iterative revisions in image and transform spaces

Area of Science:

• Image reconstruction
• Signal processing
• Computational imaging

Background:

• Limited projection data, common in medical imaging and other fields, often leads to artifacts in reconstructed images.
• Existing reconstruction techniques struggle to produce high-quality results when data is sparse or angular coverage is restricted.

Purpose of the Study:

• To develop and evaluate an iterative technique for improving image reconstruction quality from limited projection data.
• To generalize existing algorithms by incorporating a priori object information during the iterative process.

Main Methods:

• An iterative algorithm is proposed that repeatedly transforms between image and transform spaces.
• A priori object information, such as spatial boundaries and parameter variation limits, is applied at each iteration.
• The method is a generalization of the Gerchberg-Papoulis algorithm, incorporating spatial constraints into Fourier domain extrapolation.

Main Results:

• Computer simulations demonstrate the effectiveness of the iterative technique.
• Forcing the reconstructed image to conform to a priori object data significantly reduces artifacts.
• The method shows clear improvements in image quality compared to standard techniques when data is limited.

Conclusions:

• The proposed iterative technique effectively enhances image reconstruction quality with limited projection data.
• Incorporating a priori object information is crucial for mitigating artifacts and improving accuracy.
• This generalized approach offers a powerful tool for undersampled reconstruction problems.