Graduate Institute of Communication Engineering, National Taiwan University, Taipei 10617, Taiwan.
Graduate Institute of Photonics and Optoelectronics, National Taiwan University, Taipei 10617, Taiwan.
Comput Med Imaging Graph. 2021 Jan;87:101833. doi: 10.1016/j.compmedimag.2020.101833. Epub 2020 Nov 27.
Full-field optical coherence tomography (FF-OCT) has been developed to obtain three-dimensional (3D) OCT data of human skin for early diagnosis of skin cancer. Detection of dermal epidermal junction (DEJ), where melanomas and basal cell carcinomas originate, is an essential step for skin cancer diagnosis. However, most existing DEJ detection methods consider each cross-sectional frame of the 3D OCT data independently, leaving the relationship between neighboring frames unexplored. In this paper, we exploit the continuity of 3D OCT data to enhance DEJ detection. In particular, we propose a method for noise reduction of the training data and a multi-directional convolutional neural network to predict the probability of epidermal pixels in the 3D OCT data, which is more stable than one-directional convolutional neural network for DEJ detection. Our crosscheck refinement method also exploits the domain knowledge to generate a smooth DEJ surface. The average mean error of the entire DEJ detection system is approximately 6 μm.
全场光学相干断层扫描（FF-OCT）已被开发用于获取人体皮肤的三维（3D）OCT 数据，以实现皮肤癌的早期诊断。检测黑素瘤和基底细胞癌起源的表皮真皮交界处（DEJ）是皮肤癌诊断的重要步骤。然而，大多数现有的 DEJ 检测方法独立考虑 3D OCT 数据的每个横截面帧，而未探索相邻帧之间的关系。在本文中，我们利用 3D OCT 数据的连续性来增强 DEJ 检测。具体来说，我们提出了一种用于训练数据降噪的方法和一种多方向卷积神经网络，以预测 3D OCT 数据中表皮像素的概率，该方法比用于 DEJ 检测的单向卷积神经网络更稳定。我们的交叉检查细化方法还利用领域知识生成平滑的 DEJ 表面。整个 DEJ 检测系统的平均均方误差约为 6μm。
