Department of Electronics and Communication Engineering, National Institute of Technology Puducherry, Karaikal, Puducherry - 609609, India.
School of Bioengineering, VIT Bhopal University, Sehore, Madhya Pradesh - 466114, India.
Comput Biol Med. 2022 Sep;148:105852. doi: 10.1016/j.compbiomed.2022.105852. Epub 2022 Jul 12.
Highly focused images of skin captured with ordinary cameras, called macro-images, are extensively used in dermatology. Being highly focused views, the macro-images contain only lesions and background regions. Hence, the localization of lesions on the macro-images is a simple thresholding problem. However, algorithms that offer an accurate estimate of threshold and retain consistent performance on different dermatological macro-images are rare. A deep learning model, termed 'Deep Threshold Prediction Network (DTP-Net)', is proposed in this paper to address this issue. For training the model, grayscale versions of the macro-images are fed as input to the model, and the corresponding gray-level threshold values at which the Dice similarity index (DSI) between the segmented and the ground-truth images are maximized are defined as the targets. The DTP-Net exhibited the least value of root mean square error for the predicted threshold, compared with 11 state-of-the-art threshold estimation algorithms (such as Otsu's thresholding, Valley emphasized otsu's thresholding, Isodata thresholding, Histogram slope difference distribution-based thresholding, Minimum error thresholding, Poisson's distribution-based minimum error thresholding, Kapur's maximum entropy thresholding, Entropy-weighted otsu's thresholding, Minimum cross-entropy thresholding, Type-2 fuzzy-based thresholding, and Fuzzy entropy thresholding). The DTP-Net could learn the difference between the lesion and background in the intensity space and accurately predict the threshold that separates the lesion from the background. The proposed DTP-Net can be integrated into the segmentation module in automated tools that detect skin cancer from dermatological macro-images.
利用普通相机拍摄的高度聚焦的皮肤图像，称为宏观图像，在皮肤病学中得到了广泛的应用。由于高度聚焦，宏观图像仅包含病变和背景区域。因此，病变在宏观图像上的定位是一个简单的阈值问题。然而，提供准确的阈值估计并在不同的皮肤病学宏观图像上保持一致性能的算法却很少。本文提出了一种深度学习模型，称为“深度阈值预测网络（DTP-Net）”，以解决这个问题。为了训练模型，将宏观图像的灰度版本作为输入提供给模型，并且定义相应的灰度级阈值，在该阈值下分割和地面真实图像之间的骰子相似性指数（DSI）最大化。与 11 种最先进的阈值估计算法（例如 Otsu 的阈值、谷值强调的 Otsu 的阈值、Isodata 阈值、直方图斜率差分布阈值、最小误差阈值、基于泊松分布的最小误差阈值、Kapur 的最大熵阈值、基于熵权的 Otsu 的阈值、最小交叉熵阈值、基于 Type-2 模糊的阈值和模糊熵阈值）相比，DTP-Net 预测的阈值的均方根误差最小。DTP-Net 可以学习到病变和背景在强度空间中的差异，并准确预测将病变与背景分开的阈值。所提出的 DTP-Net 可以集成到自动工具的分割模块中，用于从皮肤病学宏观图像中检测皮肤癌。
