---
格式版本: 2
标题: "SPARC: Slice-to-volume Pipeline for Automated Reconstruction of gated 3D+time fetal Cardiac MRI"
原文链接: "https://arxiv.org/abs/2608.18616"
发布日期: "2026-08-19"
发布时间校准状态: "found"
发布时间需复核: "否"
发布时间来源: "rule:local:strict_original_body"
发布时间证据: "**\\[v1\\]** Wed, 19 Aug 2026 07:07:33 UTC (1,901 KB)"
发布时间校准原因: "规则确认唯一严格发布时间，来源 local:strict_original_body"
发布时间校准置信度: "high"
发布时间候选数量: 18
发布时间严格候选数量: 6
发布时间原页读取状态: "source template page reused from URL open"
发布时间未找到原因: ""
发布时间校准时间: "2026-08-20T15:23:36+08:00"
发布时间仲裁状态: "skipped"
发布时间仲裁尝试次数: 0
发布时间仲裁耗时毫秒: 0
发现时间: "2026-08-20T15:18:01+08:00"
入库时间: "2026-08-20T07:23:36.610Z"
来源平台: "arXiv 学术论文搜索"
搜索渠道: "source_template"
搜索词: "https://arxiv.org/search/?query=deployment&searchtype=all"
匹配关键词:
  - "deployment"
  - "performance"
相关厂家:
  []
相关专家:
  []
内容类型: "网页"
抓取工具: "Free Fetch + Defuddle"
清洗工具: "Defuddle Markdown + Defuddle/Readability 正文提取"
原始附件:
  []
AI优质: "否"
AI打分: 5
AI分档: "非优质"
AI质检状态: "不通过"
AI打分理由: "内容为胎儿心脏MRI重建论文，与超节点/AI Rack/机柜级AI基础设施完全无关，仅搜索词含deployment，属明显无关。"
AI质检模型: "tx-deepseek-v4-flash"
AI质检时间: "2026-08-20T15:27:44+08:00"
AI主题相关性: 0
AI来源权威性: 3
AI新颖性: 1
AI技术细节: 1
AI商业部署信号: 0
AI完整性: 0
AI摘要: "SPARC提出自动化三维时序胎儿心脏MRI重建流程，结合物理信息切片到体素重建与深度学习分割、解剖重定向，将重建时间由约49分钟降至4.8分钟；在121例临床队列中82.6%实现全自动处理，并已作为临床研究工具部署。"
AI摘要模型: "ali-deepseek-v4-flash"
AI摘要时间: "2026-09-07T03:20:11.603Z"
采集批次: "2026年8月20日14点19分32秒"
采集批次ID: "20260820-141932-079"
去重键: "https://arxiv.org/abs/2608.18616"
---

## Computer Science > Computer Vision and Pattern Recognition

## Title:SPARC: Slice-to-volume Pipeline for Automated Reconstruction of gated 3D+time fetal Cardiac MRI

[View PDF](https://arxiv.org/pdf/2608.18616) [HTML (experimental)](https://arxiv.org/html/2608.18616v1)

> Abstract:Fetal cardiac MRI (fCMR) provides valuable diagnostic information complementary to echocardiography, particularly for complex congenital heart disease (CHD). Dynamic cine imaging captures cardiac motion essential for assessment of cardiac function; however, the reconstruction of 3D+time cine volumes from 2D+time acquired slices remains challenging due to unpredictable fetal motion and the absence of automated and robust processing tools suitable for clinical deployment. We present the SPARC pipeline (Slice-to-volume Pipeline for Automated Reconstruction of gated 3D+time fetal Cardiac MRI) which combines physics-informed slice-to-volume reconstruction (SVR) of Doppler ultrasound (DUS) gated stacks of slices, assisted by deep learning (DL) models for thoracic segmentation and anatomical reorientation. The proposed SVR algorithm achieves a tenfold reduction in reconstruction time relative to existing frame-wise approaches ($4.8 \pm 1.0$ vs $49.0 \pm 14.1$ min, $p < 0.0001$) while improving the reconstruction quality. Thoracic segmentation performance using ensemble aggregation exceeded inter-rater agreement (Dice $84.7 \pm 3.9\%$ vs $81.4 \pm 7.7\%$, $p<0.05$), while anatomical reorientation achieved a success rate of $90.1\\%$. End-to-end evaluation on a large held-out clinical cohort ($n = 121$) demonstrated fully automatic processing in $82.6\\%$ of cases with a mean runtime of $7.1 \pm 1.3$ min, compatible with clinical deployment. The complete SPARC pipeline is publicly available as a Docker container [this https URL](https://hub.docker.com/r/aboutill/sparc) and is currently deployed at our institution as a clinical research tool.

| Subjects: | Computer Vision and Pattern Recognition (cs.CV) |
| --- | --- |
| Cite as: | [arXiv:2608.18616](https://arxiv.org/abs/2608.18616) \[cs.CV\] |
|  | (or [arXiv:2608.18616v1](https://arxiv.org/abs/2608.18616v1) \[cs.CV\] for this version) |
|  | [https://doi.org/10.48550/arXiv.2608.18616](https://doi.org/10.48550/arXiv.2608.18616) |

## Submission history

From: Arnaud Boutillon \[[view email](https://arxiv.org/show-email/0aabc38f/2608.18616)\]  
**\[v1\]** Wed, 19 Aug 2026 07:07:33 UTC (1,901 KB)

[Which authors of this paper are endorsers?](https://arxiv.org/auth/show-endorsers/2608.18616) | Disable MathJax ([What is MathJax?](https://info.arxiv.org/help/mathjax.html))