Decoder conditioning improves bone selective x ray image synthesis quality
RFS-UNet: Decoder-Conditioned High-Resolution Skip Recalibration for Bone-Selective DRR Synthesis
Computer Vision and Pattern Recognition
Summary
Bone-selective digitally reconstructed radiographs (DRRs) are medical images that require detailed processing to accurately show bone structures. The researchers tested if using information from the image decoder, rather than only the encoder, improves image quality. They developed a method called RFS-UNet which adjusts skip connections with both encoder and decoder data. This method slightly but consistently improved image similarity measures and reduced errors on held-out cases without increasing complexity much. The findings show that decoder information helps in refining high-resolution bone image synthesis.
What this means in practice
- •For medical image engineers: Improve bone-selective projection image synthesis in CT-based diagnostic tool pipelines using decoder-conditioned skip recalibration.
- •For radiology software developers: Refine high-resolution digitally reconstructed radiographs for better visualization and analysis of bone structures.
Authors
Xiaoyang Li, Yixuan Liu, Yuan Chai
Abstract
Bone-selective digitally reconstructed radiograph (DRR) synthesis depends on high-resolution encoder detail, yet static skips cannot condition reuse on the evolving decoder representation. We ask whether decoder state adds useful information beyond encoder-only self-recalibration for high-resolution skip reuse. RFS-UNet uses pooled encoder and aligned decoder statistics for bounded residual channel recalibration at the 512^2 and 256^2 skips, leaving the backbone unchanged. In the matched seed-2026 comparison isolating decoder conditioning, RFS raises validation PSNR by 0.254 dB over Self-RFS. Across three seeds, locked-test PSNR rises from 33.225+/-0.048 to 33.537+/-0.128 dB; RFS lowers MAE in 179/200 held-out CT cases and reduces mean MAE by 3.91%. It adds 0.117% parameters and 1.169% counted Conv2d operations. These results support decoder state as a useful conditioning signal for high-resolution feature reuse in controlled paired projection synthesis.