arXiv:2609.08044cs.CV2026-09

通过解码器状态优化高分辨率特征重用,提升骨骼选择性数字重建影像质量。

RFS-UNet: Decoder-Conditioned High-Resolution Skip Recalibration for Bone-Selective DRR Synthesis

论文配图:RFS-UNet: Decoder-Conditioned High-Resolution Skip Recalibration for Bone-Selective DRR Synthesis
图 1 · 摘自论文原文
  • 利用编码器与解码器统计信息,动态重校正高分辨率跳跃连接特征。
  • 在三个随机种子下,峰值信噪比提升0.312 dB,均方误差降低3.91%。
  • 适用于医学影像生成任务,尤其适合需要精确骨骼重建的放疗模拟场景。

骨骼选择性数字重建影像(DRR)合成依赖于高分辨率编码器细节,但静态跳跃连接无法根据不断演化的解码器表示进行条件化复用。本文探讨解码器状态是否能在高分辨率跳跃连接复用中提供超越仅编码器自重校正的有用信息。RFS-UNet 在 512² 和 256² 跳跃连接处,结合池化编码器与对齐解码器统计量,实现有界残差通道重校正,不改变主干网络结构。在匹配种子-2026 的对比实验中,相比 Self-RFS,RFS 将验证集 PSNR 提升 0.254 dB。在三个种子下的锁定测试中,PSNR 从 33.225±0.048 提升至 33.537±0.128 dB;在 200 个保留的 CT 病例中,179 例的 MAE 降低,平均 MAE 减少 3.91%。模型额外增加 0.117% 参数和 1.169% 计算量的 Conv2d 操作。结果表明,在受控配对投影合成中,解码器状态是高分辨率特征复用的有用条件信号。

原文摘要 · Abstract (English)

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.

医学影像DRR合成特征重校正

Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品;中文卡片由大模型生成,请以原文为准。