亲爱的研友该休息了!由于当前在线用户较少,发布求助请尽量完整地填写文献信息,科研通机器人24小时在线,伴您度过漫漫科研夜!身体可是革命的本钱,早点休息,好梦!

Research of using image noise characteristic to evaluation image quality in digital mammography

作者
Tianliang Kang,Jing Li
出处
期刊:Chinese journal of radiology [Chinese Medical Association]
卷期号:46 (10): 932-936
标识
DOI:10.3760/cma.j.issn.1005-1201.2012.10.014
摘要

Objective To find the proper way of assessing the image noise characteristic through the analysis of noise characteristic in the digital mammographic images.Methods The CDMAM phantom was firstly exposed with different filter materials and tube voltages in the range of 23 to 35 kVp at a fixed tube current 90 mAs,then exposed with different filter materials and tube current in the range of 20 to 200 mAs at a fixed tube voltage 29 kVp.The parameters such as background standard deviation,joint standard deviation,subtracted standard deviation and relative standard deviation were calculated and measured from both 60 pixels of signal disc and background beside of the disk for all the images.The IQFinvs were calculated for the achieved images.Non-parametric rank sum tests were used to compare the difference of the background standard deviation,joint standard deviation and subtracted standard deviation,t test was used to compare the IQFinvs between Mo and Rh filers in each group.Results With the same exposure parameters in the test above,the maximum value was the background standard deviation,while the subtracted standard deviation was the minimum,and the value of joint standard deviation was in the middle. With other parameters unchanged,the values of three kind of standard deviation increased with the increasing of tube voltage or tube current,while the differences between them getting more significant.Under the condition of 90 mAs and Mo filter material,background standard deviation,joint standard deviation,subtracted standard deviation were 10.91,10.34 and 9.87,the results were 11.24,10.57 and 9.96 while Rh filter material was used,significant difference was found ( x2 =26.0,P < 0.01 ). Under the condition of 29 kVp and Mo filter material,background standard deviation,joint standard deviation,subtracted standard deviation were 12.30,11.61 and 11.05,the results were 12.29,11.58 and 10.87 while Rh filter material was used,significant difference was found( x2 =20.0,P < 0.01 ).The IQFinvs were 149.14 ± 23.87 and 139.16 ± 23.14 under the condition of 90 mAs and Mo filter material ( t =9.129,P < 0.01 ).The IQFinvs were 149.75 ±34.85 and 145.2 ±34.64 while the parameters were 29 kVp and Rh filter material(t =1.304,P >0.05 ).The value of relative standard deviation decreased with the increasing of tube voltage or tube current,and the image quality got better with the decreasing of relative standard deviation.Conclusion Relative standard deviation is suitable not only for the expression of the digital mammographic images,but also for the evaluation of the image quality. Key words: Mammography ;  Noise;  Comparative study

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
慕月发布了新的文献求助10
15秒前
标致访旋完成签到,获得积分10
27秒前
淡然的半烟完成签到,获得积分10
35秒前
外向如雪完成签到,获得积分10
1分钟前
陶醉妙松完成签到,获得积分10
1分钟前
kobee完成签到 ,获得积分10
1分钟前
年轻火车完成签到,获得积分10
2分钟前
韶绍完成签到 ,获得积分10
2分钟前
陶醉成协完成签到,获得积分10
2分钟前
追寻孤萍完成签到,获得积分10
2分钟前
chen完成签到 ,获得积分10
2分钟前
ll完成签到 ,获得积分10
2分钟前
心灵美晓完成签到,获得积分10
3分钟前
李爱国应助wpz采纳,获得10
3分钟前
优秀函完成签到,获得积分10
3分钟前
3分钟前
Freya1528应助科研通管家采纳,获得20
3分钟前
3分钟前
wpz发布了新的文献求助10
3分钟前
李健应助wpz采纳,获得10
3分钟前
魔幻梦曼完成签到,获得积分10
3分钟前
4分钟前
wpz发布了新的文献求助10
4分钟前
满意的苑博完成签到,获得积分10
4分钟前
Jaysmith001完成签到 ,获得积分0
4分钟前
简单的不平完成签到,获得积分10
4分钟前
田様应助科研通管家采纳,获得10
5分钟前
体贴雪萍完成签到,获得积分10
5分钟前
善良的寒珊完成签到,获得积分10
5分钟前
FashionBoy应助雪山冰川采纳,获得10
5分钟前
5分钟前
雪山冰川发布了新的文献求助10
5分钟前
279完成签到,获得积分10
6分钟前
光亮静槐完成签到 ,获得积分10
6分钟前
小西西完成签到,获得积分10
6分钟前
稳重仇天完成签到,获得积分10
6分钟前
闪闪小凡完成签到,获得积分10
6分钟前
周周南完成签到 ,获得积分0
6分钟前
6分钟前
英俊的鹤完成签到,获得积分10
6分钟前
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
Rosenblum, Global Change Biology 800
自動車の空力技術 800
Essentials of Carbohydrate Chemistry and Biochemistry, 4th Edition 800
Organizational Behavior 510
Management and the Arts 510
Matrix Methods in Data Mining and Pattern Recognition Second Edition 510
热门求助领域 (近24小时)
化学 材料科学 医学 生物 纳米技术 计算机科学 化学工程 工程类 有机化学 物理 复合材料 生物化学 内科学 细胞生物学 基因 遗传学 免疫学 冶金 光电子学 癌症研究
热门帖子
关注 科研通微信公众号,转发送积分 7778271
求助须知:如何正确求助?哪些是违规求助? 9318750
关于积分的说明 20365730
捐赠科研通 7365282
什么是DOI,文献DOI怎么找? 3319178
关于科研通互助平台的介绍 2466967
邀请新用户注册赠送积分活动 2334513