用贝叶斯方法改进随时间变化的弥散核磁共振成像中的微结构估算

IF 3.3 2区医学 Q1 RADIOLOGY, NUCLEAR MEDICINE & MEDICAL IMAGING

Journal of Magnetic Resonance Imaging Pub Date : 2024-05-20 DOI:10.1002/jmri.29434

Kuiyuan Liu PhD, Zixuan Lin PhD, Tianshu Zheng PhD, Ruicheng Ba PhD, Zelin Zhang PhD, Haotian Li PhD, Hongxi Zhang MD, Assaf Tal PhD, Dan Wu PhD

{"title":"用贝叶斯方法改进随时间变化的弥散核磁共振成像中的微结构估算","authors":"Kuiyuan Liu PhD, Zixuan Lin PhD, Tianshu Zheng PhD, Ruicheng Ba PhD, Zelin Zhang PhD, Haotian Li PhD, Hongxi Zhang MD, Assaf Tal PhD, Dan Wu PhD","doi":"10.1002/jmri.29434","DOIUrl":null,"url":null,"abstract":"<div>\n \n <section>\n \n <h3> Background</h3>\n \n <p>Accurately fitting diffusion-time-dependent diffusion MRI (<i>t</i><sub><i>d</i></sub>-dMRI) models poses challenges due to complex and nonlinear formulas, signal noise, and limited clinical data acquisition.</p>\n </section>\n \n <section>\n \n <h3> Purpose</h3>\n \n <p>Introduce a Bayesian methodology to refine microstructural fitting within the IMPULSED (Imaging Microstructural Parameters Using Limited Spectrally Edited Diffusion) model and optimize the prior distribution within the Bayesian framework.</p>\n </section>\n \n <section>\n \n <h3> Study Type</h3>\n \n <p>Retrospective.</p>\n </section>\n \n <section>\n \n <h3> Population</h3>\n \n <p>Involving 69 pediatric patients (median age 6 years, interquartile range [IQR] 3–9 years, 61% male) with 41 low-grade and 28 high-grade gliomas, of which 76.8% were identified within the brainstem or cerebellum.</p>\n </section>\n \n <section>\n \n <h3> Field Strength/Sequence</h3>\n \n <p>3 T, oscillating gradient spin-echo (OGSE) and pulsed gradient spin-echo (PGSE).</p>\n </section>\n \n <section>\n \n <h3> Assessment</h3>\n \n <p>The Bayesian method's performance in fitting cell diameter (<span></span><math>\n <mrow>\n <mi>d</mi>\n </mrow></math>), intracellular volume fraction (<span></span><math>\n <mrow>\n <msub>\n <mi>f</mi>\n <mtext>in</mtext>\n </msub>\n </mrow></math>), and extracellular diffusion coefficient (<span></span><math>\n <mrow>\n <msub>\n <mi>D</mi>\n <mi>ex</mi>\n </msub>\n </mrow></math>) was compared against the NLLS method, considering simulated and experimental data. The tumor region-of-interest (ROI) were manually delineated on the <i>b</i>0 images. The diagnostic performance in distinguishing high- and low-grade gliomas was assessed, and fitting accuracy was validated against H&E-stained pathology.</p>\n </section>\n \n <section>\n \n <h3> Statistical Tests</h3>\n \n <p>T-test, receiver operating curve (ROC), area under the curve (AUC) and DeLong's test were conducted. Significance considered at <i>P</i> < 0.05.</p>\n </section>\n \n <section>\n \n <h3> Results</h3>\n \n <p>Bayesian methodology manifested increased accuracy with robust estimates in simulation (RMSE decreased by 29.6%, 40.9%, 13.6%, and STD decreased by 29.2%, 43.5%, and 24.0%, respectively for <span></span><math>\n <mrow>\n <mi>d</mi>\n </mrow></math>, <span></span><math>\n <mrow>\n <msub>\n <mi>f</mi>\n <mtext>in</mtext>\n </msub>\n </mrow></math>, and <span></span><math>\n <mrow>\n <msub>\n <mi>D</mi>\n <mi>ex</mi>\n </msub>\n </mrow></math> compared to NLLS), indicating fewer outliers and reduced error. Diagnostic performance for tumor grade was similar in both methods, however, Bayesian method generated smoother microstructural maps (outliers ratio decreased by 45.3% ± 19.4%) and a marginal enhancement in correlation with H&E staining result (<i>r</i> = 0.721 for <span></span><math>\n <mrow>\n <msub>\n <mi>f</mi>\n <mtext>in</mtext>\n </msub>\n </mrow></math> compared to <i>r</i> = 0.698 using NLLS, <i>P</i> = 0.5764).</p>\n </section>\n \n <section>\n \n <h3> Data Conclusion</h3>\n \n <p>The proposed Bayesian method substantially enhances the accuracy and robustness of IMPULSED model estimation, suggesting its potential clinical utility in characterizing cellular microstructure.</p>\n </section>\n \n <section>\n \n <h3> Evidence Level</h3>\n \n <p>3</p>\n </section>\n \n <section>\n \n <h3> Technical Efficacy</h3>\n \n <p>Stage 1</p>\n </section>\n </div>","PeriodicalId":16140,"journal":{"name":"Journal of Magnetic Resonance Imaging","volume":"61 2","pages":"724-734"},"PeriodicalIF":3.3000,"publicationDate":"2024-05-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Improving Microstructural Estimation in Time-Dependent Diffusion MRI With a Bayesian Method\",\"authors\":\"Kuiyuan Liu PhD, Zixuan Lin PhD, Tianshu Zheng PhD, Ruicheng Ba PhD, Zelin Zhang PhD, Haotian Li PhD, Hongxi Zhang MD, Assaf Tal PhD, Dan Wu PhD\",\"doi\":\"10.1002/jmri.29434\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<div>\\n \\n <section>\\n \\n <h3> Background</h3>\\n \\n <p>Accurately fitting diffusion-time-dependent diffusion MRI (<i>t</i><sub><i>d</i></sub>-dMRI) models poses challenges due to complex and nonlinear formulas, signal noise, and limited clinical data acquisition.</p>\\n </section>\\n \\n <section>\\n \\n <h3> Purpose</h3>\\n \\n <p>Introduce a Bayesian methodology to refine microstructural fitting within the IMPULSED (Imaging Microstructural Parameters Using Limited Spectrally Edited Diffusion) model and optimize the prior distribution within the Bayesian framework.</p>\\n </section>\\n \\n <section>\\n \\n <h3> Study Type</h3>\\n \\n <p>Retrospective.</p>\\n </section>\\n \\n <section>\\n \\n <h3> Population</h3>\\n \\n <p>Involving 69 pediatric patients (median age 6 years, interquartile range [IQR] 3–9 years, 61% male) with 41 low-grade and 28 high-grade gliomas, of which 76.8% were identified within the brainstem or cerebellum.</p>\\n </section>\\n \\n <section>\\n \\n <h3> Field Strength/Sequence</h3>\\n \\n <p>3 T, oscillating gradient spin-echo (OGSE) and pulsed gradient spin-echo (PGSE).</p>\\n </section>\\n \\n <section>\\n \\n <h3> Assessment</h3>\\n \\n <p>The Bayesian method's performance in fitting cell diameter (<span></span><math>\\n <mrow>\\n <mi>d</mi>\\n </mrow></math>), intracellular volume fraction (<span></span><math>\\n <mrow>\\n <msub>\\n <mi>f</mi>\\n <mtext>in</mtext>\\n </msub>\\n </mrow></math>), and extracellular diffusion coefficient (<span></span><math>\\n <mrow>\\n <msub>\\n <mi>D</mi>\\n <mi>ex</mi>\\n </msub>\\n </mrow></math>) was compared against the NLLS method, considering simulated and experimental data. The tumor region-of-interest (ROI) were manually delineated on the <i>b</i>0 images. The diagnostic performance in distinguishing high- and low-grade gliomas was assessed, and fitting accuracy was validated against H&E-stained pathology.</p>\\n </section>\\n \\n <section>\\n \\n <h3> Statistical Tests</h3>\\n \\n <p>T-test, receiver operating curve (ROC), area under the curve (AUC) and DeLong's test were conducted. Significance considered at <i>P</i> < 0.05.</p>\\n </section>\\n \\n <section>\\n \\n <h3> Results</h3>\\n \\n <p>Bayesian methodology manifested increased accuracy with robust estimates in simulation (RMSE decreased by 29.6%, 40.9%, 13.6%, and STD decreased by 29.2%, 43.5%, and 24.0%, respectively for <span></span><math>\\n <mrow>\\n <mi>d</mi>\\n </mrow></math>, <span></span><math>\\n <mrow>\\n <msub>\\n <mi>f</mi>\\n <mtext>in</mtext>\\n </msub>\\n </mrow></math>, and <span></span><math>\\n <mrow>\\n <msub>\\n <mi>D</mi>\\n <mi>ex</mi>\\n </msub>\\n </mrow></math> compared to NLLS), indicating fewer outliers and reduced error. Diagnostic performance for tumor grade was similar in both methods, however, Bayesian method generated smoother microstructural maps (outliers ratio decreased by 45.3% ± 19.4%) and a marginal enhancement in correlation with H&E staining result (<i>r</i> = 0.721 for <span></span><math>\\n <mrow>\\n <msub>\\n <mi>f</mi>\\n <mtext>in</mtext>\\n </msub>\\n </mrow></math> compared to <i>r</i> = 0.698 using NLLS, <i>P</i> = 0.5764).</p>\\n </section>\\n \\n <section>\\n \\n <h3> Data Conclusion</h3>\\n \\n <p>The proposed Bayesian method substantially enhances the accuracy and robustness of IMPULSED model estimation, suggesting its potential clinical utility in characterizing cellular microstructure.</p>\\n </section>\\n \\n <section>\\n \\n <h3> Evidence Level</h3>\\n \\n <p>3</p>\\n </section>\\n \\n <section>\\n \\n <h3> Technical Efficacy</h3>\\n \\n <p>Stage 1</p>\\n </section>\\n </div>\",\"PeriodicalId\":16140,\"journal\":{\"name\":\"Journal of Magnetic Resonance Imaging\",\"volume\":\"61 2\",\"pages\":\"724-734\"},\"PeriodicalIF\":3.3000,\"publicationDate\":\"2024-05-20\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Journal of Magnetic Resonance Imaging\",\"FirstCategoryId\":\"3\",\"ListUrlMain\":\"https://onlinelibrary.wiley.com/doi/10.1002/jmri.29434\",\"RegionNum\":2,\"RegionCategory\":\"医学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q1\",\"JCRName\":\"RADIOLOGY, NUCLEAR MEDICINE & MEDICAL IMAGING\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of Magnetic Resonance Imaging","FirstCategoryId":"3","ListUrlMain":"https://onlinelibrary.wiley.com/doi/10.1002/jmri.29434","RegionNum":2,"RegionCategory":"医学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"RADIOLOGY, NUCLEAR MEDICINE & MEDICAL IMAGING","Score":null,"Total":0}

引用次数: 0

摘要

背景：目的：介绍一种贝叶斯方法，以完善IMPULSED（使用有限光谱编辑的扩散成像微结构参数）模型中的微结构拟合，并在贝叶斯框架内优化先验分布：研究类型：回顾性研究：研究对象：69名儿童患者（中位年龄6岁，四分位距[IQR]3-9岁，61%为男性），其中41例为低级别胶质瘤，28例为高级别胶质瘤，76.8%的胶质瘤位于脑干或小脑：3T、振荡梯度自旋回波（OGSE）和脉冲梯度自旋回波（PGSE）：贝叶斯方法在拟合细胞直径（d $$ d $$）、细胞内体积分数（f in $$ {f}_{in} $$）和细胞外扩散系数（D ex $$ {D}_{ex} $$）方面的性能与 NLLS 方法进行了比较，并考虑了模拟和实验数据。肿瘤感兴趣区（ROI）是在 b0 图像上人工划定的。评估了区分高级别和低级别胶质瘤的诊断性能，并根据 H&E 染色病理学验证了拟合的准确性：进行了 T 检验、接收者操作曲线（ROC）、曲线下面积（AUC）和 DeLong 检验。结果：贝叶斯方法提高了准确性：贝叶斯方法在模拟中表现出更高的准确性和稳健的估计值（与 NLLS 相比，d $$ d$ 、f in $$ {f}_{in} $$ 和 D ex $$ {D}_{ex} $$ 的 RMSE 分别降低了 29.6%、40.9%、13.6%，STD 分别降低了 29.2%、43.5% 和 24.0%），表明异常值更少，误差更小。两种方法对肿瘤分级的诊断性能相似，但贝叶斯方法生成的显微结构图更平滑（异常值比例降低了 45.3% ± 19.4%），与 H&E 染色结果的相关性略有增强（f in $$ {f}_{in} $$ 的 r = 0.721，而使用 NLLS 的 r = 0.698，P = 0.5764）：数据结论：所提出的贝叶斯方法大大提高了 IMPULSED 模型估计的准确性和稳健性，表明该方法在表征细胞微观结构方面具有潜在的临床实用性。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

查看原文本刊更多论文

Improving Microstructural Estimation in Time-Dependent Diffusion MRI With a Bayesian Method

Background

Accurately fitting diffusion-time-dependent diffusion MRI (t_d-dMRI) models poses challenges due to complex and nonlinear formulas, signal noise, and limited clinical data acquisition.

Purpose

Introduce a Bayesian methodology to refine microstructural fitting within the IMPULSED (Imaging Microstructural Parameters Using Limited Spectrally Edited Diffusion) model and optimize the prior distribution within the Bayesian framework.

Study Type

Retrospective.

Population

Involving 69 pediatric patients (median age 6 years, interquartile range [IQR] 3–9 years, 61% male) with 41 low-grade and 28 high-grade gliomas, of which 76.8% were identified within the brainstem or cerebellum.

Field Strength/Sequence

3 T, oscillating gradient spin-echo (OGSE) and pulsed gradient spin-echo (PGSE).

Assessment

The Bayesian method's performance in fitting cell diameter ( $d$ ), intracellular volume fraction ( $f_{in}$ ), and extracellular diffusion coefficient ( $D_{ex}$ ) was compared against the NLLS method, considering simulated and experimental data. The tumor region-of-interest (ROI) were manually delineated on the b0 images. The diagnostic performance in distinguishing high- and low-grade gliomas was assessed, and fitting accuracy was validated against H&E-stained pathology.

Statistical Tests

T-test, receiver operating curve (ROC), area under the curve (AUC) and DeLong's test were conducted. Significance considered at P < 0.05.

Results

Bayesian methodology manifested increased accuracy with robust estimates in simulation (RMSE decreased by 29.6%, 40.9%, 13.6%, and STD decreased by 29.2%, 43.5%, and 24.0%, respectively for $d$ , $f_{in}$ , and $D_{ex}$ compared to NLLS), indicating fewer outliers and reduced error. Diagnostic performance for tumor grade was similar in both methods, however, Bayesian method generated smoother microstructural maps (outliers ratio decreased by 45.3% ± 19.4%) and a marginal enhancement in correlation with H&E staining result (r = 0.721 for $f_{in}$ compared to r = 0.698 using NLLS, P = 0.5764).

Data Conclusion

The proposed Bayesian method substantially enhances the accuracy and robustness of IMPULSED model estimation, suggesting its potential clinical utility in characterizing cellular microstructure.

Evidence Level

Technical Efficacy

Stage 1

求助全文

通过发布文献求助，成功后即可免费获取论文全文。去求助

来源期刊

Journal of Magnetic Resonance Imaging 医学-核医学

CiteScore

9.70

自引率

6.80%

发文量

494

审稿时长

2 months

期刊介绍： The Journal of Magnetic Resonance Imaging (JMRI) is an international journal devoted to the timely publication of basic and clinical research, educational and review articles, and other information related to the diagnostic applications of magnetic resonance.