Pathologists need to combine information from differently stained pathological slices to obtain accurate diagnostic results. Deformable image registration is a necessary technique for fusing multi-modal pathological slices. This paper proposes a hybrid deep feature-based deformable image registration framework for stained pathological samples. We first extract dense feature points and perform points matching by two deep learning feature networks. Then, to further reduce false matches, an outlier detection method combining the isolation forest statistical model and the local affine correction model is proposed. Finally, the interpolation method generates the DVF for pathology image registration based on the above matching points. We evaluate our method on the dataset of the Non-rigid Histology Image Registration (ANHIR) challenge, which is co-organized with the IEEE ISBI 2019 conference. Our technique outperforms the traditional approaches by 17% with the Average-Average registration target error (rTRE) reaching 0.0034. The proposed method achieved state-of-the-art performance and ranking it 1 in evaluating the test dataset. The proposed hybrid deep feature-based registration method can potentially become a reliable method for pathology image registration.

翻译：病理学家需要综合来自不同污染病理切片的信息,以获得准确的诊断结果。 变形图像登记是使用多模式病理切片的一种必要技术。 本文建议为染色病理样本建立一个混合的基于深度地貌的变形图像登记框架。 我们首先提取稠密的特征点, 并通过两个深层学习特征网络进行匹配。 然后, 为进一步减少假匹配, 提出了一个将孤立森林统计模型与本地缝合校正模型相结合的异常检测方法。 最后, 套用方法产生了基于上述匹配点的病理图像登记DVF。 我们评估了我们关于非硬性地貌图象登记数据集的方法, 与 IEEEE ISBI 2019 会议共同组织。 我们的技术比传统方法高出17%, 与平均地貌登记目标错误( rTRE) 达到0.0034。 拟议的方法在评估测试数据集时达到了最先进的性能和排序。 拟议的深层混合地貌登记方法有可能成为可靠的图像登记路径法。