Yahyatabar Mohammad, Jouvet Philippe, Cheriet Farida
Annu Int Conf IEEE Eng Med Biol Soc. 2020 Jul;2020:1242-1245. doi: 10.1109/EMBC44109.2020.9176033.
Automatic and accurate lung segmentation in chest X-ray (CXR) images is fundamental for computer-aided diagnosis systems since the lung is the region of interest in many diseases and also it can reveal useful information by its contours. While deep learning models have reached high performances in the segmentation of anatomical structures, the large number of training parameters is a concern since it increases memory usage and reduces the generalization of the model. To address this, a deep CNN model called Dense-Unet is proposed in which, by dense connectivity between various layers, information flow increases throughout the network. This lets us design a network with significantly fewer parameters while keeping the segmentation robust. To the best of our knowledge, Dense-Unet is the lightest deep model proposed for the segmentation of lung fields in CXR images. The model is evaluated on the JSRT and Montgomery datasets and experiments show that the performance of the proposed model is comparable with state-of-the-art methods.
在胸部X光（CXR）图像中实现自动且准确的肺部分割，对于计算机辅助诊断系统而言至关重要，因为肺部是许多疾病的关注区域，而且其轮廓也能揭示有用信息。虽然深度学习模型在解剖结构分割方面已取得了高性能，但大量的训练参数令人担忧，因为这会增加内存使用并降低模型的泛化能力。为解决这一问题，提出了一种名为Dense-Unet的深度卷积神经网络（CNN）模型，通过各层之间的密集连接，信息流在整个网络中得以增加。这使我们能够设计出一个参数显著更少的网络，同时保持分割的稳健性。据我们所知，Dense-Unet是为CXR图像中的肺野分割所提出的最轻量级深度模型。该模型在JSRT和蒙哥马利数据集上进行了评估，实验表明所提模型的性能与当前最先进的方法相当。
