Abstract:Purpose: Low-dose CT has attracted increasing attention due to growing concerns about radiation exposure in medical scans. However, the frugal use of x-ray radiation inevitably reduces the quality of the CT images, introducing artifacts such as noise and streaks which make the reconstructed images difficult to read in clinical routine. For follow-up CT exams a prior scan is often available. It typically contains the same anatomical structures, just somewhat deformed and not aligned. This work describes a two-step technique that utilizes this prior scan to achieve high-quality low-dose CT imaging, overcoming difficulties arising from noise artifacts and misalignment. We specifically focus on reducing the dose by lowering the number of projections. This gives rise to severe streak artifacts which possibly lower the readability of CT images to a larger extent than the fine-grained noise that results from lowering the mA or kV settings. Methods: A common approach is to apply image filtering to reduce the noise artifacts. These techniques typically utilize pixel neighborhoods in the degraded image to estimate the true value of a pixel at the center of this neighborhood. However, this can lead to poor results when the image is severely contaminated under very low low-dose situations. We propose a method that utilizes the nondegraded, clean prior to determine higher quality pixel statistics to form the pixel estimates, supported by the matching scheme of the non-local means filter. To make this matching reliable, a good registration of prior and low-dose image is required. For this, we employ a state-of-the-art registration method, called SIFT-flow, which can tolerate the high amount of streak noise. But even for properly registered images, using an artifact free prior for the matching yields inferior results. We hence describe a scheme that first constructs a tandem-prior with streak artifacts resembling those in the low-dose image, and then employs this image for the matching, but uses the corresponding high-quality prior to determine the pixel estimates. Results: Two experimental studies are carried out, using a head phantom and a human lung with projections gathered via simulation. We assess the quality of the processed reconstruction with various metrics: mathematical and perceptual. We find that the quality that can be obtained with the artifact-matched prior-based scheme significantly exceeds that of all competing schemes. Even though the general prior-based approach is able to eliminate the streak artifacts, only the artifact-matched scheme can restore small detail and feature sharpness. Conclusions: The reduced-projection low-dose image reconstruction algorithm we present outperforms traditional image restoration algorithms when a prior scan is available. Our method is quite efficient and as such it is well suited for fast-paced clinical applications such as image-assisted interventions, orthopedic alignment scans, and follow-ups.

Prior-image-based low-dose CT reconstruction for adaptive radiation therapy

Full Dose Ct Database Induced Reconstruction With Nonlocal Means Prior For Ultra-Low-Dose Lung Ct: A Preliminary Study

Redundancy Information Induced Edge-Preserving Prior for Perfusion CT Image Reconstruction

Synthetic CT generation from cone-beam CT using deep-learning for breast adaptive radiotherapy

Low-dose CT Reconstruction Method Based on Prior Information of Normal-Dose Image

Study of peripheral dose from low-dose CT to adaptive radiotherapy of postoperative prostate cancer

Deep learning-based low-dose CT for adaptive radiotherapy of abdominal and pelvic tumors

Deep Learning-Based Low Dose Ct Imaging

Low-Dose CT Reconstruction Using Dataset-free Learning

A Self-supervised Deep Learning Network for Low-Dose CT Reconstruction

Low dose CT reconstruction assisted by an image manifold prior

Efficient low-dose CT artifact mitigation using an artifact-matched prior scan

DaNet: dose-aware network embedded with dose-level estimation for low-dose CT imaging

Combining Deep Learning and Adaptive Sparse Modeling for Low-dose CT Reconstruction

Re-UNet: A Novel Multi-scale Reverse U-shaped Network Architecture for Low-dose CT Image Reconstruction

AHP-Net: adaptive-hyper-parameter deep learning based image reconstruction method for multilevel low-dose CT

AI-Enabled Ultra-Low-Dose CT Reconstruction

Deep Radon Prior: A Fully Unsupervised Framework for Sparse-View CT Reconstruction

Low-dose Lung CT Image Restoration Using Adaptive Prior Features from Full-dose Training Database.

Iterative reconstruction of low-dose CT based on differential sparse

Low dose CT reconstruction with nonlocal means-based prior predicted from normal-dose CT database.