Uncertainty for Safe Utilization of Machine Learning in Medical Imaging - 4th International Workshop, UNSURE 2022, Held in Conjunction with MICCAI 2022, Singapore, September 18, 2022, Proceedings
Uncertainty for Safe Utilization of Machine Learning in Medical Imaging - 4th International Workshop, UNSURE 2022, Held in Conjunction with MICCAI 2022, Singapore, September 18, 2022, Proceedings
复制标题
医学影像中机器学习安全利用的不确定性 - 第四届国际研讨会，UNSURE 2022，与 MICCAI 2022 联合举行，新加坡，2022 年 9 月 18 日，会议记录
DOI：
10.1007/978-3-031-16749-2_6
复制
发表时间：
2022
期刊：
影响因子：
--
通讯作者：
Ouyang C
中科院分区：
文献类型：
--
作者：
Ouyang C
深度模型的概率校准在诸如医学成像的安全关键应用中是非常期望的。它通过将预测概率与测试数据中的实际准确度相匹配，使深度网络的输出概率变得可解释。在图像分割中，良好校准的概率允许放射科医生识别模型预测分割不可靠的区域。这些不可靠的预测通常发生在由成像伪影或看不见的成像协议引起的域外（OOD）图像中。不幸的是，大多数以前的图像分割的校准方法执行次优OOD图像。为了减少OOD图像时的校准误差，我们提出了一种新的事后校准模型。我们的模型利用像素的敏感性对扰动在局部水平上，在全球范围内的形状先验信息。该模型在心脏MRI分割数据集上进行测试，这些数据集包含看不见的成像伪影和来自看不见的成像协议的图像。我们证明减少校准误差相比，国家的最先进的校准算法。
Probability calibration for deep models is highly desirable in safety-critical applications such as medical imaging. It makes output probabilities of deep networks interpretable, by aligning prediction probability with the actual accuracy in test data. In image segmentation, well-calibrated probabilities allow radiologists to identify regions where model-predicted segmentations are unreliable. These unreliable predictions often occur to out-of-domain (OOD) images that are caused by imaging artifacts or unseen imaging protocols. Unfortunately, most previous calibration methods for image segmentation perform sub-optimally on OOD images. To reduce the calibration error when confronted with OOD images, we propose a novel post-hoc calibration model. Our model leverages the pixel susceptibility against perturbations at the local level, and the shape prior information at the global level. The model is tested on cardiac MRI segmentation datasets that contain unseen imaging artifacts and images from an unseen imaging protocol. We demonstrate reduced calibration errors compared with the state-of-the-art calibration algorithm.