Exploring uncertainty measures in deep networks for Multiple sclerosis lesion detection and segmentation

体素 分割 辍学(神经网络) 人工智能 计算机科学 深度学习 背景(考古学) 病变 模式识别(心理学) 图像分割 机器学习 医学 病理 生物 古生物学
作者
T. R. Gopalakrishnan Nair,Doina Precup,Douglas L. Arnold,Tal Arbel
出处
期刊:Medical Image Analysis [Elsevier]
卷期号:59: 101557-101557 被引量:307
标识
DOI:10.1016/j.media.2019.101557
摘要

Deep learning networks have recently been shown to outperform other segmentation methods on various public, medical-image challenge datasets, particularly on metrics focused on large pathologies. For diseases such as Multiple Sclerosis (MS), however, monitoring all the focal lesions visible on MRI sequences, even very small ones, is essential for disease staging, prognosis, and evaluating treatment efficacy. Small lesion segmentation presents significant challenges to popular deep learning models. This, coupled with their deterministic predictions, hinders their clinical adoption. Uncertainty estimates for these predictions would permit subsequent revision by clinicians. We present the first exploration of multiple uncertainty estimates based on Monte Carlo (MC) dropout (Gal and Ghahramani, 2016) in the context of deep networks for lesion detection and segmentation in medical images. Specifically, we develop a 3D MS lesion segmentation CNN, augmented to provide four different voxel-based uncertainty measures based on MC dropout. We train the network on a proprietary, large-scale, multi-site, multi-scanner, clinical MS dataset, and compute lesion-wise uncertainties by accumulating evidence from voxel-wise uncertainties within detected lesions. We analyze the performance of voxel-based segmentation and lesion-level detection by choosing operating points based on the uncertainty. Uncertainty filtering improves both voxel and lesion-wise TPR and FDR on remaining, certain predictions compared to sigmoid-based TPR/FDR curves. Small lesions and lesion-boundaries are the most uncertain regions, which is consistent with human-rater variability.
最长约 10秒,即可获得该文献文件

科研通智能强力驱动
Strongly Powered by AbleSci AI
更新
大幅提高文件上传限制,最高150M (2024-4-1)

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
啊啊啊橙子完成签到,获得积分20
1秒前
等待的靖雁完成签到 ,获得积分10
1秒前
1秒前
DumBell完成签到,获得积分10
2秒前
az发布了新的文献求助10
2秒前
特特特特特雷西完成签到,获得积分10
3秒前
XL神放发布了新的文献求助10
3秒前
3秒前
完美世界应助辛坦夫采纳,获得10
4秒前
www完成签到,获得积分10
5秒前
zhang完成签到,获得积分10
6秒前
兔子给兔子的求助进行了留言
7秒前
7秒前
7秒前
领导范儿应助kingmantj采纳,获得10
8秒前
8秒前
一路硕博应助wdy采纳,获得10
8秒前
隐形曼青应助腼腆的缘分采纳,获得10
8秒前
9秒前
9秒前
9秒前
9秒前
li完成签到,获得积分10
10秒前
ding应助小胡采纳,获得20
11秒前
tuanheqi应助Linda采纳,获得50
11秒前
Ariel发布了新的文献求助20
12秒前
12秒前
12秒前
感动含雁完成签到,获得积分10
12秒前
强健的鼠标完成签到,获得积分10
12秒前
LINGO完成签到,获得积分10
13秒前
cannon8发布了新的文献求助10
13秒前
tina3058完成签到,获得积分10
14秒前
樱桃汽水怪兽完成签到,获得积分10
15秒前
16秒前
16秒前
jim发布了新的文献求助10
17秒前
18秒前
科研通AI2S应助漂亮的灯泡采纳,获得10
19秒前
20秒前
高分求助中
Sustainability in ’Tides Chemistry 2000
The ACS Guide to Scholarly Communication 2000
Studien zur Ideengeschichte der Gesetzgebung 1000
TM 5-855-1(Fundamentals of protective design for conventional weapons) 1000
Threaded Harmony: A Sustainable Approach to Fashion 810
Pharmacogenomics: Applications to Patient Care, Third Edition 800
Gerard de Lairesse : an artist between stage and studio 500
热门求助领域 (近24小时)
化学 医学 生物 材料科学 工程类 有机化学 生物化学 物理 内科学 纳米技术 计算机科学 化学工程 复合材料 基因 遗传学 催化作用 物理化学 免疫学 量子力学 细胞生物学
热门帖子
关注 科研通微信公众号,转发送积分 3074970
求助须知:如何正确求助?哪些是违规求助? 2728319
关于积分的说明 7503288
捐赠科研通 2376446
什么是DOI,文献DOI怎么找? 1260052
科研通“疑难数据库(出版商)”最低求助积分说明 610789
版权声明 597117