MATLAB-based Parallel Beam Filtered Back Projection Reconstruction Algorithm

In this article, we introduce the MATLAB-based parallel beam filtered back projection reconstruction algorithm. This algorithm is primarily used in medical imaging applications, where it creates three-dimensional images by projecting X-ray beams from different angles onto an object and measuring the transmission rates. The implementation typically involves key MATLAB functions such as radon() for projection data acquisition and iradon() for the inverse Radon transform reconstruction. The algorithm has numerous applications, particularly in tumor detection where physicians can utilize this method to determine tumor size and location accurately. Additionally, this algorithm finds significant utility in industrial sectors for detecting material defects or determining internal structures of machine components. The filtered back projection process involves two main stages: first applying a convolution filter (such as Ram-Lak, Shepp-Logan, or Cosine filters) to the projection data to reduce blurring artifacts, followed by back projection to reconstruct the image. The MATLAB implementation typically includes steps for projection data preprocessing, frequency domain filtering using fft() and ifft() functions, and back projection using interpolation techniques. The algorithm's efficiency can be enhanced through vectorization and optimizing the back projection calculations. In summary, the MATLAB-based parallel beam filtered back projection reconstruction algorithm serves as a powerful tool with extensive applications across various fields, providing reliable reconstruction results while maintaining computational efficiency through MATLAB's optimized numerical computing capabilities.