Facultad de Ingeniería, Universidad Tecnologica de Bolivar, Cartagena, Colombia.
VISILAB, Universidad de Castilla-La Mancha, E.T.S. Ingeniería Industrial, Avda Camilo Jose Cela, Ciudad Real, Spain.
PLoS One. 2024 Nov 12;19(11):e0311849. doi: 10.1371/journal.pone.0311849. eCollection 2024.
Image segmentation of the corneal endothelium with deep convolutional neural networks (CNN) is challenging due to the scarcity of expert-annotated data. This work proposes a data augmentation technique via warping to enhance the performance of semi-supervised training of CNNs for accurate segmentation. We use a unique augmentation process for images and masks involving keypoint extraction, Delaunay triangulation, local affine transformations, and mask refinement. This approach accurately captures the natural variability of the corneal endothelium, enriching the dataset with realistic and diverse images. The proposed method achieved an increase in the mean intersection over union (mIoU) and Dice coefficient (DC) metrics of 17.2% and 4.8% respectively, for the segmentation task in corneal endothelial images on multiple CNN architectures. Our data augmentation strategy successfully models the natural variability in corneal endothelial images, thereby enhancing the performance and generalization capabilities of semi-supervised CNNs in medical image cell segmentation tasks.
由于缺乏专家标注数据，使用深度卷积神经网络（CNN）对角膜内皮进行图像分割具有挑战性。本研究提出了一种通过变形进行数据增强的技术，以增强 CNN 的半监督训练性能，实现准确的分割。我们使用一种独特的图像和掩模增强过程，涉及关键点提取、Delaunay 三角剖分、局部仿射变换和掩模细化。这种方法可以准确地捕捉到角膜内皮的自然变化，通过真实多样的图像丰富数据集。该方法在多个 CNN 架构上的角膜内皮图像分割任务中，平均交并比（mIoU）和骰子系数（DC）的度量分别提高了 17.2%和 4.8%。我们的数据增强策略成功地对角膜内皮图像的自然变化进行建模，从而提高了半监督 CNN 在医学图像细胞分割任务中的性能和泛化能力。
