文献速递脑血管病

基于DWI和FLAIR生成急性缺血性卒中合成MR灌注图的深度学习模型:开发与外部验证

Generating Synthetic MR Perfusion Maps From DWI and FLAIR in Acute Ischemic Stroke: Development and External Validation of a Deep Learning Model.
PubMed 文献 · Stroke · 2026-06-18

本研究回顾性收集2个欧洲中心355例前循环卒中患者的急性期MRI数据,训练去噪扩散概率模型和生成对抗网络,从DWI、FLAIR及梗死核心mask生成合成T-max灌注图。最佳模型为2.5维去噪扩散概率模型,合成图与真实T-max>6s区域在内外部验证中Dice分别为0.82和0.59,错配分布接近真实。该无创框架有望减少对比剂灌注依赖。

要点

  • 最佳模型为2.5维去噪扩散概率模型,输入DWI、FLAIR和梗死核心mask。
  • 合成T-max图与真实图在110秒内高度相似,T-max>6s区域内部Dice 0.82、外部0.59。
  • 梗死核心mask对模型性能起关键作用,合成图可捕捉主要灌注模式。

核心结论

  • 可从非对比MRI(DWI、FLAIR及梗死核心mask)生成合成T-max灌注图,减少对比增强采集需求。

研究设计与样本

  • 回顾性研究,收集2个欧洲中心355例前循环卒中患者的急性MRI数据(海德堡2010-2018,波尔多2021-2022)。
  • 训练6个版本的去噪扩散概率模型和生成对抗网络,输入DWI、FLAIR和梗死核心mask生成合成T-max图。
  • 采用图像相似度指标比较合成与真实T-max图,并用Dice系数比较T-max>6s区域,分析错配体积分布,并行消融研究。

主要发现

  • 最佳性能来自2.5维去噪扩散概率模型,使用DWI、FLAIR、梗死核心mask和灌注加权损失函数。
  • 合成T-max图在110秒内与真实图高度相似。
  • T-max>6s区域内部验证平均Dice为0.82(SD 0.08),外部验证平均Dice为0.59(SD 0.13)。
  • 合成图与真实错配分布密切匹配,能捕捉关键灌注模式。
  • 梗死核心mask对模型性能起关键作用,DWI和FLAIR输入亦重要。

关键数值与推荐

  • T-max>6s区域内部验证平均Dice 0.82(SD 0.08),外部验证平均Dice 0.59(SD 0.13)。
  • 合成T-max图与真实图在110秒内高度相似。

临床怎么用

  • 该无创框架有望扩大急性卒中灌注数据的可及性,缩短成像流程。
  • 通过消除对比增强采集需求,可能加速治疗决策。

局限性

  • 回顾性研究设计。
  • 外部验证Dice(0.59)低于内部验证(0.82),泛化性能有待提升。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

原文

BACKGROUND: Magnetic resonance imaging (MRI) is critical for acute stroke triage, but it is time-consuming and often requires contrast injection for perfusion imaging. This study aimed to synthesize T-map perfusion maps from routinely available, noncontrast diffusion-weighted imaging and fluid-attenuated inversion recovery using deep generative models. We hypothesized that relevant perfusion information could be inferred from these modalities to streamline imaging and reduce reliance on dynamic susceptibility contrast perfusion. METHODS: Acute magnetic resonance imaging data from 355 patients with anterior circulation stroke, including dynamic susceptibility contrast perfusion, were retrospectively collected from 2 European centers (Heidelberg: 2010-2018; Bordeaux: 2021-2022). Six versions of a denoising diffusion probabilistic model and a generative adversarial network architecture were trained to generate synthetic time-to-maximum (T-max) perfusion maps from diffusion-weighted imaging, fluid-attenuated inversion recovery, and infarct core mask as inputs. Performance was assessed by comparing synthetic and ground-truth T-max maps using image similarity metrics. Regions with T-max >6 s were compared using Dice coefficients, and mismatch volume distributions were analyzed. An ablation study quantified the contribution of each input. RESULTS: The best performance was achieved by a denoising diffusion probabilistic model with a 2.5-dimensional architecture using diffusion-weighted imaging, fluid-attenuated inversion recovery, infarct core mask, and a perfusion-weighted loss function. It produced synthetic perfusion T-max maps with high similarity to ground truth under 110 s. The model showed strong spatial overlap for T-max> 6 s regions in internal validation (average Dice, 0.82; SD, 0.08) and external validation (average Dice, 0.59; SD, 0.13), respectively. Synthetic maps closely matched ground truth mismatch distributions, capturing key perfusion patterns. The infarct core mask played a critical role in model performance, alongside diffusion-weighted imaging and fluid-attenuated inversion recovery inputs. CONCLUSIONS: We propose a noninvasive, scalable framework to generate synthetic T-max perfusion maps from noncontrast magnetic resonance imaging. This approach could expand access to perfusion data in acute stroke, shorten imaging protocols, and accelerate treatment decisions by eliminating the need for contrast-enhanced acquisition.

来源:Stroke,查看原文。版权归原作者所有。

引用

  1. Matsulevits A, Koch A, Mahé-Verdure C, et al. Generating Synthetic MR Perfusion Maps From DWI and FLAIR in Acute Ischemic Stroke: Development and External Validation of a Deep Learning Model. Stroke. 2026. PMID: 42312380. doi:10.1161/STROKEAHA.126.055813

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。