Abstract:Background: In recent years, deep learning strategies have been combined with either the filtered backprojection or iterative methods or the direct projection-to-image by deep learning only to reconstruct images. Some of these methods can be applied to address the interior reconstruction problems for centered regions of interest (ROIs) with fixed sizes. Developing a method to enable interior tomography with arbitrarily located ROIs with nearly arbitrary ROI sizes inside a scanning field of view (FOV) remains an open question. Purpose: To develop a new pathway to enable interior tomographic reconstruction for arbitrarily located ROIs with arbitrary sizes using a single trained deep neural network model. Methods: The method consists of two steps. First, an analytical weighted backprojection reconstruction algorithm was developed to perform domain transform from divergent fan-beam projection data to an intermediate image feature space, B ( x ⃗ ) $B(\vec{x})$ , for an arbitrary size ROI at an arbitrary location inside the FOV. Second, a supervised learning technique was developed to train a deep neural network architecture to perform deconvolution to obtain the true image f ( x ⃗ ) $f(\vec{x})$ from the new feature space B ( x ⃗ ) $B(\vec{x})$ . This two-step method is referred to as Deep-Interior for convenience. Both numerical simulations and experimental studies were performed to validate the proposed Deep-Interior method. Results: The results showed that ROIs as small as a diameter of 5 cm could be accurately reconstructed (similarity index 0.985 ± 0.018 on internal testing data and 0.940 ± 0.025 on external testing data) at arbitrary locations within an imaging object covering a wide variety of anatomical structures of different body parts. Besides, ROIs of arbitrary size can be reconstructed by stitching small ROIs without additional training. Conclusion: The developed Deep-Interior framework can enable interior tomographic reconstruction from divergent fan-beam projections for short-scan and super-short-scan acquisitions for small ROIs (with a diameter larger than 5 cm) at an arbitrary location inside the scanning FOV with high quantitative reconstruction accuracy.

A Feature Refinement Approach for Statistical Interior Ct Reconstruction

Reweighted Anisotropic Total Variation Minimization for Limited-Angle CT Reconstruction

Constrained TV-minimization Reconstruction from Exterior CT Data

Statistical Interior Tomography via L1 Norm Dictionary Learning without Assuming an Object Support

Pet Image Reconstruction From Under-Sampled Data

An improved statistical iterative algorithm for sparse-view and limited-angle CT image reconstruction

A Fast Regularized Iterative Algorithm for Fan-Beam CT Reconstruction.

Low-dose CT reconstruction via edge-preserving total variation regularization

High Order Total Variation Method For Interior Tomography

A Modified Interval Subdividing Based Geometric Calibration Method for Interior Tomography

GPU-based Low Dose CT Reconstruction via Edge-preserving Total Variation Regularization

Deep-Interior: A new pathway to interior tomographic image reconstruction via a weighted backprojection and deep learning

Low dose PET reconstruction with total variation regularization.

A Constrained, Total-Variation Minimization Algorithm for Low-Intensity X-Ray CT.

Positron Emission Tomography Image Reconstruction Using Feature Extraction

A Reweighted Total Variation Minimization Method for Few View CT Reconstruction in the Instant CT

A Preliminary Investigation of Using Prior Information for Potentially Improving Image Reconstruction in Few-View Ct

3D Anisotropic Total Variation Method for Limited-Angle Ct Reconstruction

Preliminary Study of Image Reconstruction from Sparse-View Data in Phase-Contrast CT

Evaluation of low-dose computed tomography reconstruction using spatial-radon domain total generalized variation regularization

Deep Learning Interior Tomography for Region-of-Interest Reconstruction