[论文解读] An Analysis on Ensemble Learning optimized Medical Image Classification with Deep Convolutional Neural Networks

Dominik Müller, Iñaki Soto‐Rey+1 more|arXiv (Cornell University)|Jan 27, 2022

Radiomics and Machine Learning in Medical Imaging被引用 5

一句话总结

本文提出了一种可复现的深度学习流水线，用于医学图像分类，评估了集成学习技术——增强、堆叠和袋装——在九种CNN架构和四个医学影像数据集上的表现。堆叠方法取得了最高的性能提升，F1得分最高提升13%；而简单的池化函数表现优于复杂的基于学习的函数。

ABSTRACT

Novel and high-performance medical image classification pipelines are heavily utilizing ensemble learning strategies. The idea of ensemble learning is to assemble diverse models or multiple predictions and, thus, boost prediction performance. However, it is still an open question to what extent as well as which ensemble learning strategies are beneficial in deep learning based medical image classification pipelines. In this work, we proposed a reproducible medical image classification pipeline for analyzing the performance impact of the following ensemble learning techniques: Augmenting, Stacking, and Bagging. The pipeline consists of state-of-the-art preprocessing and image augmentation methods as well as 9 deep convolution neural network architectures. It was applied on four popular medical imaging datasets with varying complexity. Furthermore, 12 pooling functions for combining multiple predictions were analyzed, ranging from simple statistical functions like unweighted averaging up to more complex learning-based functions like support vector machines. Our results revealed that Stacking achieved the largest performance gain of up to 13% F1-score increase. Augmenting showed consistent improvement capabilities by up to 4% and is also applicable to single model based pipelines. Cross-validation based Bagging demonstrated significant performance gain close to Stacking, which resulted in an F1-score increase up to +11%. Furthermore, we demonstrated that simple statistical pooling functions are equal or often even better than more complex pooling functions. We concluded that the integration of ensemble learning techniques is a powerful method for any medical image classification pipeline to improve robustness and boost performance.

研究动机与目标

• 探究不同集成学习策略在基于深度学习的医学图像分类流水线中的性能影响。
• 评估增强、堆叠和袋装技术在多样化医学影像数据集上的有效性。
• 对比12种池化函数——从无加权平均到基于SVM的方法——以整合多个模型的预测结果。
• 确定在该场景下，复杂池化函数是否优于简单的统计函数。
• 提供一个可复现的流水线，用于医学图像分类中集成方法的基准测试。

提出的方法

• 流水线整合了最先进的医学图像预处理和数据增强技术。
• 在四个复杂度各异的公开医学影像数据集中，使用了九种预训练的深度卷积神经网络架构作为基学习器。
• 系统性地应用了三种集成学习策略：增强（通过数据增强实现模型集成）、堆叠（元学习器结合基模型预测）和袋装（基于交叉验证的模型平均）。
• 评估了12种池化函数以整合预测结果，范围从无加权平均到支持向量机。
• 使用F1得分衡量性能，并通过严格的交叉验证确保结果的稳健性和可复现性。
• 整个流水线已开源，并附带补充材料，以实现完全可复现性。

实验结果

研究问题

• RQ1 在医学图像分类中，哪种集成学习策略——增强、堆叠或袋装——能带来最高的性能提升？
• RQ2 从简单平均到复杂学习方法的各类池化函数，如何影响集成预测性能？
• RQ3 集成学习在多样化医学影像数据集上，能在多大程度上提升鲁棒性和泛化能力？
• RQ4 在该场景下，简单的统计池化函数是否能优于更复杂的基于学习的池化函数？
• RQ5 集成方法带来的性能增益是否在不同CNN架构和数据集复杂度下保持一致？

主要发现

• 堆叠方法实现了最大的性能提升，与单个模型相比，F1得分最高提升13%。
• 增强方法始终提升了性能，最高达4%，即使在单模型流水线中也有效。
• 基于交叉验证的袋装方法带来了显著的性能提升，F1得分最高提升11%，接近堆叠方法的增益。
• 简单的统计池化函数，如无加权平均，与更复杂的基于学习的池化函数（如SVM）表现相当或更优。
• 所提出的流水线在所有四个医学影像数据集中均表现出一致的性能提升，无论数据集复杂度如何。
• 集成学习的整合显著增强了医学图像分类任务中模型的鲁棒性和预测可靠性。