A fully automatic image analysis system for computer assisted diagnosis of acute ischemic stroke based on magnetic resonance diffusion-weighted imaging
Qingmao Hu; Peiyi Gao; Zhijun Chen; Li Ma; Xiaohua Du; Jianhuang Wu; Fucang Jia; Hui Chen
2009
会议名称北美放射学会(RSNA 2009)Proceedings of RSNA 95th Annual Meeting 2009
会议地点Qingdao, China
英文摘要AISSuite takes the DWI and derived apparent diffusion map (ADC) as input without PWI.  Infarct regions (IRs) are approximated by adaptive image segmentation with low ADC and high DWI, and transition regions (TRs) as neighboring IRs with intermediate ADC. For the two regions with the largest IR and sum of IR and TR, the radial distance and the average ADC gradient within the TR are approximated and their product is taken as δA; if both δAs are greater than 0.18 and at least one is larger than 0.21, then this patient has SIT; for patients with SIT, the spatial distribution of ADC within the IRs is explored to assess the suitability for TS. 45 ischemic patients from Siemens 3.0T and 2 patients from GE 1.5T retrieved within 6h of symptom onset were tested. Taking the mismatch as gold standard for existence of SIT, AISSuite can yield a sensitivity of (35/39) 89.7% and specificity of (6/8) 75%. AISSuite could exclude the 2 TS cases with SICH and recommend TS for all the cases having good recovery through TS in 1 minute.
收录类别其他
语种英语
内容类型会议论文
源URL[http://ir.siat.ac.cn:8080/handle/172644/2431]  
专题深圳先进技术研究院_集成所
作者单位2009
推荐引用方式
GB/T 7714
Qingmao Hu,Peiyi Gao,Zhijun Chen,et al. A fully automatic image analysis system for computer assisted diagnosis of acute ischemic stroke based on magnetic resonance diffusion-weighted imaging[C]. 见:北美放射学会(RSNA 2009)Proceedings of RSNA 95th Annual Meeting 2009. Qingdao, China.
个性服务
查看访问统计
相关权益政策
暂无数据
收藏/分享
所有评论 (0)
暂无评论
 

除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。


©版权所有 ©2017 CSpace - Powered by CSpace