SinoSynth: A Physics-based Domain Randomization Approach for Generalizable CBCT Image Enhancement

TL;DR

SinoSynth introduces a physics-based degradation model to generate diverse synthetic CBCT images, improving artifact removal and image quality across varied datasets without extensive real data collection.

Contribution

The paper presents a novel physics-based degradation model for CBCT, enabling effective training of generative networks with synthetic data that generalize well across multiple institutions.

Findings

Synthetic data improves artifact removal in CBCT images.

Models trained on SinoSynth data outperform those trained on real data.

The degradation model preserves anatomical structures in synthetic images.

Abstract

Cone Beam Computed Tomography (CBCT) finds diverse applications in medicine. Ensuring high image quality in CBCT scans is essential for accurate diagnosis and treatment delivery. Yet, the susceptibility of CBCT images to noise and artifacts undermines both their usefulness and reliability. Existing methods typically address CBCT artifacts through image-to-image translation approaches. These methods, however, are limited by the artifact types present in the training data, which may not cover the complete spectrum of CBCT degradations stemming from variations in imaging protocols. Gathering additional data to encompass all possible scenarios can often pose a challenge. To address this, we present SinoSynth, a physics-based degradation model that simulates various CBCT-specific artifacts to generate a diverse set of synthetic CBCT images from high-quality CT images without requiring pre-aligned data. Through extensive experiments, we demonstrate that several different generative networks trained on our synthesized data achieve remarkable results on heterogeneous multi-institutional datasets, outperforming even the same networks trained on actual data. We further show that our degradation model conveniently provides an avenue to enforce anatomical constraints in conditional generative models, yielding high-quality and structure-preserving synthetic CT images.