Wang Xun, Shi Xin, Meng Xiangyu, Zhang Zhiyuan, Zhang Chaogang
Department of Computer Science and Technology, China University of Petroleum, Qingdao, Shandong, China.
High Performance Computer Research Center, University of Chinese Academy of Sciences, Beijing, China.
Front Pharmacol. 2023 Aug 1;14:1084155. doi: 10.3389/fphar.2023.1084155. eCollection 2023.
Partially supervised learning (PSL) is urgently necessary to explore to construct an efficient universal lesion detection (ULD) segmentation model. An annotated dataset is crucial but hard to acquire because of too many Computed tomography (CT) images and the lack of professionals in computer-aided detection/diagnosis (CADe/CADx). To address this problem, we propose a novel loss function to reduce the proportion of negative anchors which is extremely likely to classify the lesion area (positive samples) as a negative bounding box, further leading to an unexpected performance. Before calculating loss, we generate a mask to intentionally choose fewer negative anchors which will backward wrongful loss to the network. During the process of loss calculation, we set a parameter to reduce the proportion of negative samples, and it significantly reduces the adverse effect of misclassification on the model. Our experiments are implemented in a 3D framework by feeding a partially annotated dataset named DeepLesion, a large-scale public dataset for universal lesion detection from CT. We implement a lot of experiments to choose the most suitable parameter, and the result shows that the proposed method has greatly improved the performance of a ULD detector. Our code can be obtained at https://github.com/PLuld0/PLuldl.
探索构建高效的通用病变检测（ULD）分割模型，部分监督学习（PSL）是迫切需要的。带注释的数据集至关重要，但由于计算机断层扫描（CT）图像数量众多且缺乏计算机辅助检测/诊断（CADe/CADx）方面的专业人员，获取起来很困难。为了解决这个问题，我们提出了一种新颖的损失函数，以减少极有可能将病变区域（正样本）分类为负边界框的负锚点比例，进而导致意外的性能表现。在计算损失之前，我们生成一个掩码，有意选择较少的负锚点，这些负锚点会将错误的损失反向传播到网络。在损失计算过程中，我们设置一个参数来减少负样本的比例，这显著降低了错误分类对模型的不利影响。我们的实验在一个3D框架中进行，通过输入一个名为DeepLesion的部分注释数据集，这是一个用于从CT进行通用病变检测的大规模公共数据集。我们进行了大量实验以选择最合适的参数，结果表明所提出的方法极大地提高了ULD检测器的性能。我们的代码可在https://github.com/PLuld0/PLuldl获取。
