How Does Pruning Impact Long-Tailed Multi-Label Medical Image Classifiers?-猫眼课题宝

DOI：

10.48550/arxiv.2308.09180

发表时间：

2023-08

期刊：

ArXiv

影响因子：

通讯作者：

中科院分区：

其他

文献类型：

作者：

G. Holste;Ziyu Jiang;A. Jaiswal;M. Hanna;Shlomo Minkowitz;A. Legasto;J. Escalon;Sharon Steinberger;M. Bittman;Thomas C. Shen;Ying Ding;R. M. Summers;G. Shih;Yifan Peng;Zhangyang Wang

剪枝已成为一种强大的技术，可用于压缩深度神经网络、减少内存使用和推理时间，而不会显着影响整体性能。然而，修剪影响模型行为的细微差别尚不清楚，特别是对于临床环境中常见的长尾、多标签数据集。在部署经过修剪的模型进行诊断时，这种知识差距可能会产生危险的影响，其中意外的模型行为可能会影响患者的健康。为了填补这一空白，我们首次分析了修剪对经过训练以通过胸部 X 光 (CXR) 诊断胸部疾病的神经网络的影响。在两个大型 CXR 数据集上，我们检查了哪些疾病受修剪影响最大，并根据疾病频率和共现行为描述了“遗忘性”类别。此外，我们还识别出未压缩模型和经过严格修剪的模型不一致的单个 CXR，称为修剪识别范例 (PIE)，并进行人类读者研究来评估它们的统一质量。我们发现放射科医生认为 PIE 具有更多的标签噪声、较低的图像质量和较高的诊断难度。这项工作代表了理解剪枝对深度长尾、多标签医学图像分类中模型行为的影响的第一步。所有代码、模型权重和数据访问说明都可以在 https://github.com/VITA-Group/PruneCXR 找到。

Pruning has emerged as a powerful technique for compressing deep neural networks, reducing memory usage and inference time without significantly affecting overall performance. However, the nuanced ways in which pruning impacts model behavior are not well understood, particularly for long-tailed, multi-label datasets commonly found in clinical settings. This knowledge gap could have dangerous implications when deploying a pruned model for diagnosis, where unexpected model behavior could impact patient well-being. To fill this gap, we perform the first analysis of pruning’s effect on neural networks trained to diagnose thorax diseases from chest X-rays (CXRs). On two large CXR datasets, we examine which diseases are most affected by pruning and characterize class “forgettability” based on disease frequency and co-occurrence behavior. Further, we identify individual CXRs where uncompressed and heavily pruned models disagree, known as pruning-identified exemplars (PIEs), and conduct a human reader study to evaluate their unifying qualities. We find that radiologists perceive PIEs as having more label noise, lower image quality, and higher diagnosis difficulty. This work represents a first step toward understanding the impact of pruning on model behavior in deep long-tailed, multi-label medical image classification. All code, model weights, and data access instructions can be found at https://github.com/VITA-Group/PruneCXR.

How Does Pruning Impact Long-Tailed Multi-Label Medical Image Classifiers?

文献摘要

相似文献