arXiv cs.AI· Qinghui Liu, Jon Andr\'e Ottesen, Thu Nguyen, Geir Ringstad, Ingrid Mossige, Siri Fl{\o}gstad Svensson, Atle Bj{\o}rnerud, Per Kristian Eide, Kyrre Eeg Emblem·· 3 天前AI 评分21
从平扫 MRI 预测脑内示踪剂分布：直接回归与流匹配的对比
Forecasting Brain Tracer Distribution from Pre-Contrast MRI: A Comparison of Direct Regression and Flow Matching
AI 导读
研究用 104 例训练患者（878 个 MRI 体积）比较直接图像到图像回归模型与基于条件流匹配的生成模型，预测鞘内注射示踪剂在脑内的后期分布。直接模型在 23 例留出患者和 51 例特发性正常压力脑积水（INPH）队列中，将"直接复制平扫图像"的误差分别降低约 60% 和 25%，且预测变化图与观察到的示踪剂强化一致；生成模型在两个队列中精度均更低。
正文 · 原文
Abstract:Intrathecal contrast-enhanced MRI shows how a tracer injected into the cerebrospinal fluid spreads through the brain, but requires repeated scans over 24--48\,h. We therefore ask how much of the later tracer distribution can be predicted from the pre-contrast scan and elapsed time alone. Using 104 training patients (878 MRI volumes), we compared a direct image-to-image regression model with a more complex generative model based on conditional flow matching, and tested both on 23 held-out patients and on 51 patients from a separate cohort with idiopathic normal pressure hydrocephalus (INPH). The direct model reduced the error of simply copying the pre-contrast scan by about 60\% in held-out patients and 25\% in the INPH cohort, and its predicted change maps agreed with the observed tracer enhancement rather than only reproducing anatomy. The generative model was less accurate in both cohorts. A substantial part of the later tracer distribution is therefore predictable from the patient's brain anatomy as seen on pre-contrast MRI, and the added model complexity gave no benefit at this data scale. With prospective validation, such forecasts could be studied for patient stratification and treatment planning in intrathecal therapies.
| Comments: | 31 pages, 13 figures, 6 tables |
| Subjects: | Image and Video Processing (eess.IV); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG) |
| Cite as: | arXiv:2609.22377 [eess.IV] |
| (or arXiv:2609.22377v2 [eess.IV] for this version) | |
| https://doi.org/10.48550/arXiv.2609.22377 arXiv-issued DOI via DataCite |
Submission history
From: Qinghui Liu [view email]
[v1]
Thu, 17 Sep 2026 20:49:59 UTC (10,692 KB)
[v2]
Tue, 13 Oct 2026 09:40:23 UTC (11,682 KB)
来源：arXiv cs.AI · arxiv.org