In recent years, deep learning models have revolutionized medical image interpretation, offering substantial improvements in diagnostic accuracy. However, these models often struggle with challenging images where critical features are partially or fully occluded, which is a common scenario in clinical practice. In this paper, we propose a novel curriculum learning-based approach to train deep learning models to handle occluded medical images effectively. Our method progressively introduces occlusion, starting from clear, unobstructed images and gradually moving to images with increasing occlusion levels. This ordered learning process, akin to human learning, allows the model to first grasp simple, discernable patterns and subsequently build upon this knowledge to understand more complicated, occluded scenarios. Furthermore, we present three novel occlusion synthesis methods, namely Wasserstein Curriculum Learning (WCL), Information Adaptive Learning (IAL), and Geodesic Curriculum Learning (GCL). Our extensive experiments on diverse medical image datasets demonstrate substantial improvements in model robustness and diagnostic accuracy over conventional training methodologies.

翻译：近年来，深度学习模型革新了医学影像解读，显著提升了诊断准确性。然而，在临床实践中，这些模型常难以处理关键特征被部分或完全遮挡的挑战性图像。本文提出一种基于课程学习的新方法，旨在有效训练深度学习模型处理遮挡医学影像。该方法渐进式引入遮挡，从清晰无遮挡的图像开始，逐步过渡到遮挡程度递增的图像。这种有序的学习过程类似于人类学习机制，使模型先掌握简单可辨的模式，进而基于此逐步理解更复杂的遮挡场景。此外，我们提出了三种新颖的遮挡合成方法：瓦瑟斯坦课程学习（WCL）、信息自适应学习（IAL）和测地课程学习（GCL）。在多样化医学图像数据集上的大量实验表明，与常规训练方法相比，该方法显著提升了模型的鲁棒性和诊断准确性。