P4‐068: Automated quantification of white matter hyperintensities to support dementia diagnosis

作者
Katherine R. Gray,Natalie A. Royle,Robin Wolz,David Hill
出处
期刊:Alzheimers & Dementia [Wiley]
卷期号:11 (7S_Part_17)
标识
DOI:10.1016/j.jalz.2015.06.1773
摘要

White matter hyperintensities (WMH) are present in various dementia types, but are particularly associated with vascular dementia (VaD). The NINDS-AIREN criteria, used for clinical diagnosis of VaD, suggest a WMH load of over 25% of total white matter as a diagnostic criterion. Automated WMH assessment methods have been applied to a range of patient groups, and the load and location of WMH provide useful information for both dementia diagnosis, and differentiation between dementia types. To bring quantitative tools into clinical practice, the data must be placed in context, and we present an automated tool that provides a summary report based on quantitative assessment of global and regional WMH. Automated quantification of WMH is performed using T1W MRI and FLAIR images, based on an in-house implementation of the “LST: Lesion Segmentation Tool” (http://www.applied-statistics.de/lst.html). The percentage of WMH within total white matter is computed, as well as the percentage of WMH within four regions of interest (juxtacortical, periventricular, deep, and juxtacortical). We performed preliminary validation of the method in 43 subjects with clinically diagnosed dementia. In the patient report, we indicate that the patient meets clinical imaging criteria for VaD if over 25% of total white matter constitutes WMH. Based on analysis of test data for which visually rated Fazekas scores were available, we determined that 5-25% WMH indicates a likely micro-vascular component of symptoms, and below 5% WMH that symptoms are unlikely to be caused by micro-vascular disease. Validation tests revealed dice overlap between manual and automated segmentations of 0.46. This is comparable with other state-of-the-art automated methods. Good correlation coefficients between manual and automated WMH volumes were also found, R = 0.81 for total WMH, R = 0.86 for juxtaventricular WMH, R = 0.84 for periventricular WMH, R = 0.78 for deep WMH, and R = 0.81 for juxtacortical WMH.

科研通智能强力驱动
Strongly Powered by AbleSci AI
科研通是完全免费的文献互助平台,具备全网最快的应助速度,最高的求助完成率。 对每一个文献求助,科研通都将尽心尽力,给求助人一个满意的交代。
实时播报
任性豪英发布了新的文献求助10
刚刚
刚刚
烟花应助颜谨采纳,获得10
1秒前
小慕发布了新的文献求助10
1秒前
明亮书芹发布了新的文献求助10
1秒前
zbearupz发布了新的文献求助10
2秒前
Hello应助陈鑫采纳,获得10
2秒前
大个应助芝麻小丸子采纳,获得10
2秒前
太阳完成签到,获得积分10
3秒前
华仔应助小张采纳,获得30
3秒前
3秒前
4秒前
知之发布了新的文献求助10
4秒前
4秒前
5秒前
喜悦的琳完成签到,获得积分10
5秒前
5秒前
Jasper应助正直若颜采纳,获得10
5秒前
地平线完成签到,获得积分10
5秒前
果子发布了新的文献求助10
6秒前
斯文元正应助忐忑的访彤采纳,获得10
6秒前
6秒前
6秒前
6秒前
7秒前
7秒前
8秒前
8秒前
8秒前
Minigun完成签到,获得积分10
9秒前
9秒前
三声完成签到 ,获得积分10
9秒前
传奇3应助1234采纳,获得10
9秒前
吴小根发布了新的文献求助10
9秒前
Liiiii发布了新的文献求助10
10秒前
roryt应助文林宝贝采纳,获得10
10秒前
10秒前
Shann发布了新的文献求助10
10秒前
从容书雁发布了新的文献求助10
10秒前
10秒前
高分求助中
(应助此贴封号)【重要!!请各用户(尤其是新用户)详细阅读】【科研通的精品贴汇总】 10000
Römisch-Germanische Forschungen 1000
China Pluperfect I: Epistemology of Past and Outside in Chinese Art 520
Matrix Methods in Data Mining and Pattern Recognition Second Edition 510
The fast track to determining transfer functions of linear circuits: The student guide 500
The Analytical and Numerical Solution of Electric and Magnetic Fields 500
Green Fire Retardants for Polymeric Materials 400
热门求助领域 (近24小时)
化学 材料科学 医学 生物 纳米技术 工程类 有机化学 化学工程 生物化学 计算机科学 内科学 物理 复合材料 催化作用 细胞生物学 无机化学 光电子学 物理化学 电极 基因
热门帖子
关注 科研通微信公众号,转发送积分 7616156
求助须知:如何正确求助?哪些是违规求助? 9191586
关于积分的说明 19696718
捐赠科研通 7188778
什么是DOI,文献DOI怎么找? 3271575
关于科研通互助平台的介绍 2434637
邀请新用户注册赠送积分活动 2266740