Bayesian deep learning outperforms clinical trial estimators of intracerebral and intraventricular hemorrhage volume.
Bayesian deep learning outperforms clinical trial estimators of intracerebral and intraventricular hemorrhage volume.
复制标题
DOI：
10.1111/jon.12997
复制
发表时间：
2022-09
期刊：
影响因子：
--
通讯作者：
中科院分区：
文献类型：
--
作者：
Intracerebral hemorrhage (ICH) and intraventricular hemorrhage (IVH) clinical trials rely on manual linear and semi-quantitative (LSQ) estimators like the ABC/2, modified Graeb and IVH scores for timely volumetric estimation from CT. Deep learning (DL) volumetrics of ICH have recently approached the accuracy of gold-standard planimetry. However, DL and LSQ strategies have been limited by unquantified uncertainty, in particular when ICH and IVH estimates intersect. Bayesian deep learning methods can be used to approximate uncertainty, presenting an opportunity to improve quality assurance in clinical trials. A DL model was trained to simultaneously segment ICH and IVH using diagnostic CT data from the Minimally Invasive Surgery Plus Alteplase for ICH Evacuation (MISTIE) III and Clot Lysis: Evaluating Accelerated Resolution of IVH (CLEAR) III clinical trials. Bayesian uncertainty approximation was performed using Monte-Carlo dropout. We compared the performance of our model with estimators used in the CLEAR IVH and MISTIE II trials. The reliability of planimetry, DL and LSQ volumetrics in the setting of high ICH and IVH intersection is quantified using consensus estimates. Our DL model volume correlations and median Dice scores of 0.994 and 0.946 for ICH in MISTIE II, and 0.980 and 0.863 for IVH in CLEAR IVH respectively, outperforming LSQ estimates from the clinical trials. We found significant linear relationships between ICH uncertainty, Dice scores (r=−0.849) and relative volume difference (r=0.735). In our validation clinical trial dataset, DL models with Bayesian uncertainty approximation provided superior volumetric estimates to LSQ methods with real-time estimates of model uncertainty.
登录
查看更多内容
脑部出血疏散（Mistie）中微创手术加上高度的安全性和功效：一项随机，控制，开放标签，第2期试验。
影响因子：
48
作者：
Hanley, Daniel F.;Thompson, Richard E.;Muschelli, John;Rosenblum, Michael;McBee, Nichol;Lane, Karen;Bistran-Hall, Amanda J.;Mayo, Steven W.;Keyl, Penelope;Gandhi, Dheeraj;Morgan, Tim C.;Ullman, Natalie;Mould, W. Andrew;Carhuapoma, J. Ricardo;Kase, Carlos;Ziai, Wendy;Thompson, Carol B.;Yenokyan, Gayane;Huang, Emily;Broaddus, William C.;Graham, R. Scott;Aldrich, E. Francois;Dodd, Robert;Wijman, Cristanne;Caron, Jean-Louis;Huang, Judy;Camarata, Paul;Mendelow, A. David;Gregson, Barbara;Janis, Scott;Vespa, Paul;Martin, Neil;Awad, Issam;Zuccarello, Mario
通讯作者：
Zuccarello, Mario
影响因子：
8.3
作者：
Hemphill, JC;Bonovich, DC;Johnston, SC
通讯作者：
Johnston, SC
影响因子：
168.9
作者：
Hanley, Daniel F.;Thompson, Richard E.;Awad, Issam A.
通讯作者：
Awad, Issam A.
影响因子：
158.5
作者：
Mayer, Stephan A.;Brun, Nikolai C.;Steiner, Thorsten
通讯作者：
Steiner, Thorsten
DOI：
10.1016/s0140-6736(16)32410-2
发表时间：
2017-02-11
期刊：
Lancet (London, England)
影响因子：
--
作者：
Hanley DF;Lane K;McBee N;Ziai W;Tuhrim S;Lees KR;Dawson J;Gandhi D;Ullman N;Mould WA;Mayo SW;Mendelow AD;Gregson B;Butcher K;Vespa P;Wright DW;Kase CS;Carhuapoma JR;Keyl PM;Diener-West M;Muschelli J;Betz JF;Thompson CB;Sugar EA;Yenokyan G;Janis S;John S;Harnof S;Lopez GA;Aldrich EF;Harrigan MR;Ansari S;Jallo J;Caron JL;LeDoux D;Adeoye O;Zuccarello M;Adams HP Jr;Rosenblum M;Thompson RE;Awad IA;CLEAR III Investigators
通讯作者：
CLEAR III Investigators