论文标题
USCL：通过视频对比表示学习预处理深度超声图像诊断模型
USCL: Pretraining Deep Ultrasound Image Diagnosis Model through Video Contrastive Representation Learning
论文作者
论文摘要
基于大多数深神经网络（DNNS）的超声（US）医学图像分析模型使用验证的骨架（例如ImageNet）进行更好的模型概括。但是，自然图像和医学图像之间的域间隙会导致不可避免的性能瓶颈。为了减轻这个问题，为在同一域上直接预修的是一个名为us-4的美国数据集。它包含来自美国四个视频子数据库的23,000张图像。为了从US-4中学习强大的功能，我们提出了一种美国半监督的对比学习方法，名为USCL，以预处理。为了避免负面对之间的高相似之处以及有限的美国视频中的大量视觉特征，USCL采用了样本对生成方法来丰富与单一的对比优化步骤相关的功能。在几个下游任务上进行了广泛的实验，表明了USCL预处理对Imagenet预处理和其他最先进的（SOTA）预处理的优势。特别是，USCL预处理的主链在Pocus数据集上实现了超过94％的微调精度，该准确性比Imagenet预算模型的84％高10％。这项工作的源代码可在https://github.com/983632847/USCL上获得。
Most deep neural networks (DNNs) based ultrasound (US) medical image analysis models use pretrained backbones (e.g., ImageNet) for better model generalization. However, the domain gap between natural and medical images causes an inevitable performance bottleneck. To alleviate this problem, an US dataset named US-4 is constructed for direct pretraining on the same domain. It contains over 23,000 images from four US video sub-datasets. To learn robust features from US-4, we propose an US semi-supervised contrastive learning method, named USCL, for pretraining. In order to avoid high similarities between negative pairs as well as mine abundant visual features from limited US videos, USCL adopts a sample pair generation method to enrich the feature involved in a single step of contrastive optimization. Extensive experiments on several downstream tasks show the superiority of USCL pretraining against ImageNet pretraining and other state-of-the-art (SOTA) pretraining approaches. In particular, USCL pretrained backbone achieves fine-tuning accuracy of over 94% on POCUS dataset, which is 10% higher than 84% of the ImageNet pretrained model. The source codes of this work are available at https://github.com/983632847/USCL.
