Deep-Learning-Based Segmentation of the Shoulder from MRI with Inference Accuracy Prediction.
Deep-Learning-Based Segmentation of the Shoulder from MRI with Inference Accuracy Prediction.
复制标题
从MRI进行深度学习的分割，并推断精度预测。
DOI：
10.3390/diagnostics13101668
复制
发表时间：
2023-05-09
期刊：
影响因子：
3.6
通讯作者：
Gerber, Kate
中科院分区：
文献类型：
--
作者：
Hess, Hanspeter;Ruckli, Adrian C.;Burki, Finn;Gerber, Nicolas;Menzemer, Jennifer;Burger, Juergen;Schar, Michael;Zumstein, Matthias A.;Gerber, Kate
基于三维（3D）图像的肩袖撕裂患者解剖分析已被提出作为一种改善修复预后分析的方法，以减少术后再撕裂的发生率。然而，对于临床应用，需要一种有效且鲁棒的方法用于从MRI分割解剖结构。我们介绍了使用深度学习网络自动分割肱骨、肩胛骨和肩袖肌肉，并集成自动结果验证。在19家临床试验机构采集的76例肩袖撕裂患者的N = 111例诊断性T1加权MRI上进行训练，并在N = 60例诊断性T1加权MRI上进行测试，nnU-Net对解剖结构进行分割，平均Dice系数为0.91 ± 0.06。为了在推理过程中自动识别不准确的分割，nnU-Net框架被调整为允许直接从其子网络估计标签特定的网络不确定性。分割的平均Dice系数由子网络识别的需要分割校正的标签产生，平均灵敏度为1.0，特异性为0.94。所提出的自动方法通过消除对耗时的手动分割和逐切片分割验证的需要，促进了在临床常规中使用3D诊断。
Three-dimensional (3D)-image-based anatomical analysis of rotator cuff tear patients has been proposed as a way to improve repair prognosis analysis to reduce the incidence of postoperative retear. However, for application in clinics, an efficient and robust method for the segmentation of anatomy from MRI is required. We present the use of a deep learning network for automatic segmentation of the humerus, scapula, and rotator cuff muscles with integrated automatic result verification. Trained on N = 111 and tested on N = 60 diagnostic T1-weighted MRI of 76 rotator cuff tear patients acquired from 19 centers, a nnU-Net segmented the anatomy with an average Dice coefficient of 0.91 ± 0.06. For the automatic identification of inaccurate segmentations during the inference procedure, the nnU-Net framework was adapted to allow for the estimation of label-specific network uncertainty directly from its subnetworks. The average Dice coefficient of segmentation results from the subnetworks identified labels requiring segmentation correction with an average sensitivity of 1.0 and a specificity of 0.94. The presented automatic methods facilitate the use of 3D diagnosis in clinical routine by eliminating the need for time-consuming manual segmentation and slice-by-slice segmentation verification.
登录
查看更多内容
DOI：
10.1148/ryai.220132
发表时间：
2023-03-01
期刊：
RADIOLOGY-ARTIFICIAL INTELLIGENCE
影响因子：
--
作者：
Riem, Lara;Feng, Xue;Blemker, Silvia S.
通讯作者：
Blemker, Silvia S.
影响因子：
2
作者：
Zeng G;Schmaranzer F;Degonda C;Gerber N;Gerber K;Tannast M;Burger J;Siebenrock KA;Zheng G;Lerch TD
通讯作者：
Lerch TD
影响因子：
4.6
作者：
Ro K;Kim JY;Park H;Cho BH;Kim IY;Shim SB;Choi IY;Yoo JC
通讯作者：
Yoo JC
影响因子：
5.7
作者：
Roy, Abhijit Guha;Conjeti, Sailesh;Wachinger, Christian
通讯作者：
Wachinger, Christian
影响因子：
2.3
作者：
Longo UG;Carnevale A;Piergentili I;Berton A;Candela V;Schena E;Denaro V
通讯作者：
Denaro V