Reliability of Automated Intracranial Volume Measurements by Synthetic Brain MRI in Children

Author:

Weiss Veronika1,Vishwanathan Nathan12,Dutschke Anja1,Stranger Nikolaus1ORCID,Scherkl Mario1,Nagy Eszter1ORCID,Ciornei-Hoffman Andreea13ORCID,Tschauner Sebastian1ORCID

Affiliation:

1. Division of Pediatric Radiology, Department of Radiology, Medical University of Graz, 8036 Graz, Austria

2. Second Faculty of Medicine, Charles University, 116 36 Prague, Czech Republic

3. Department of Radiology and Medical Imaging, ”Iuliu Hatieganu” University of Medicine and Pharmacy, 400347 Cluj-Napoca, Romania

Abstract

(1) Background: Hydrocephalus poses challenges in pediatric neuroimaging, and conventional MRI methods have limitations regarding its accurate quantification. Synthetic MRI (SyMRI) offers a promising automated solution to assess intracranial compartment volumes. However, its clinical utility in pediatric patients remains underexplored. Our study aims to assess the accuracy and reliability of automated CSF volume measurements using SyMRI in children and adolescents, comparing them with manual measurements and human expert ratings. (2) Methods: A single-center retrospective study included 124 pediatric patients undergoing cranial MRI with SyMRI. CSF, brain parenchyma, and intracranial volumes were measured using both automated SyMRI and manual methods. Human radiologists assessed hydrocephalus subjectively. (3) Results: Correlations between manual and SyMRI volume evaluations were significant. Human raters demonstrated good agreement on hydrocephalus ratings among themselves (Fleiss’ kappa = 0.66, p < 0.001) but only moderate agreement with the SyMRI method (Cohen’s kappa = 0.45, p < 0.001). SyMRI volumes were systematically tendentially higher in SyMRI (CSF p = 0.005; BPV and ICV p < 0.001). (4) Conclusions: Our findings highlight SyMRI’s reliability in assessing hydrocephalus and intracranial volumes in pediatric cases. Despite some differences from manual measurements, the strong correlation suggests its clinical viability.

Publisher

MDPI AG

同舟云学术

1.学者识别学者识别

2.学术分析学术分析

3.人才评估人才评估

"同舟云学术"是以全球学者为主线,采集、加工和组织学术论文而形成的新型学术文献查询和分析系统,可以对全球学者进行文献检索和人才价值评估。用户可以通过关注某些学科领域的顶尖人物而持续追踪该领域的学科进展和研究前沿。经过近期的数据扩容,当前同舟云学术共收录了国内外主流学术期刊6万余种,收集的期刊论文及会议论文总量共计约1.5亿篇,并以每天添加12000余篇中外论文的速度递增。我们也可以为用户提供个性化、定制化的学者数据。欢迎来电咨询!咨询电话:010-8811{复制后删除}0370

www.globalauthorid.com

TOP

Copyright © 2019-2024 北京同舟云网络信息技术有限公司
京公网安备11010802033243号  京ICP备18003416号-3