DOMINO: Domain-Aware Model Calibration in Medical Image Segmentation-猫眼课题宝

DOI：

10.1007/978-3-031-16443-9_44

发表时间：

2022-09

期刊：

ArXiv

影响因子：

通讯作者：

中科院分区：

其他

文献类型：

作者：

Skylar E. Stolte;Kyle Volle;A. Indahlastari;Alejandro Albizu;A. Woods;Kevin Brink;Matthew Hale;R. Fang

模型校准测量预测概率估计值与真实正确性可能性之间的一致性。正确的模型校准对于高风险应用至关重要。不幸的是，现代深度神经网络的校准很差，影响了可信度和可靠性。由于组织边界的自然不确定性，医学图像分割特别受到这种影响。这是激怒了他们的损失函数，这有利于过度自信的大多数类。我们解决这些挑战与DOMINO，域感知模型校准方法，利用类标签之间的语义混淆性和层次相似性。我们的实验表明，我们的DOMINO校准深度神经网络在头部图像分割方面的性能优于未校准模型和最先进的形态测量方法。我们的研究结果表明，我们的方法可以始终实现更好的校准，更高的准确性，更快的推理时间比这些方法，特别是在罕见的类。这种性能归功于我们的域感知正则化，以通知语义模型校准。这些发现表明了类标签之间的语义联系在建立深度学习模型的信心方面的重要性。该框架有可能提高通用医学图像分割模型的可信度和可靠性。这篇文章的代码可以在https://github.com/lab-smile/DOMINO上找到。

Model calibration measures the agreement between the predicted probability estimates and the true correctness likelihood. Proper model calibration is vital for high-risk applications. Unfortunately, modern deep neural networks are poorly calibrated, compromising trustworthiness and reliability. Medical image segmentation particularly suffers from this due to the natural uncertainty of tissue boundaries. This is exasperated by their loss functions, which favor overconfidence in the majority classes. We address these challenges with DOMINO, a domain-aware model calibration method that leverages the semantic confusability and hierarchical similarity between class labels. Our experiments demonstrate that our DOMINO-calibrated deep neural networks outperform non-calibrated models and state-of-the-art morphometric methods in head image segmentation. Our results show that our method can consistently achieve better calibration, higher accuracy, and faster inference times than these methods, especially on rarer classes. This performance is attributed to our domain-aware regularization to inform semantic model calibration. These findings show the importance of semantic ties between class labels in building confidence in deep learning models. The framework has the potential to improve the trustworthiness and reliability of generic medical image segmentation models. The code for this article is available at: https://github.com/lab-smile/DOMINO.

DOMINO: Domain-Aware Model Calibration in Medical Image Segmentation

文献摘要

相似文献