Univ. Grenoble Alpes, CNRS, CHU Grenoble Alpes, Grenoble INP, TIMC-IMAG, Grenoble, France.
Med Phys. 2022 Aug;49(8):5138-5148. doi: 10.1002/mp.15679. Epub 2022 Apr 29.
Prostate segmentation of 3D TRUS images is a prerequisite for several diagnostic and therapeutic applications. Unfortunately, this difficult task suffers from high intra and interobserver variability, even for experienced urologists/radiologists. This is why automatic segmentation algorithms could have a significant clinical added-value.
This paper introduces a new deep segmentation architecture consisting of two main phases: view-specific segmentations of 2D slices and their fusion. The segmentation phase is based on three segmentation networks trained in parallel on specific slice viewing directions: axial, coronal, and sagittal. The proposed fusion network is then fed with the output of the segmentation networks and trained to produce three confidence maps. These maps correspond to the local trust granted by the fusion network to each view-specific segmentation network. Finally, for a given slice, the segmentation is computed by combining these confidence maps with their corresponding segmentations. The 3D segmentation of the prostate is obtained by restacking all the segmented slices to form a volume.
This approach was evaluated on a database of 100 patients with several combinations of network architectures (for both the segmentation phase and the fusion phase) to show the flexibility and reliability of the framework. The proposed approach was also compared to STAPLE, to the majority voting strategy, and to a direct 3D approach tested on the same database. The new method outperforms these three approaches on all evaluation criteria. Finally, the results of the multi-eXpert fusion (MXF) framework compare favorably with other state-of-the-art methods, while these methods typically work on smaller databases.
We proposed a novel MXF framework to segment 3D TRUS images of the prostate. The main feature of this approach is the fusion of expert network results at the pixel level using computed confidence maps. Experiments conducted on a clinical database have shown the robustness and flexibility of this approach and its superiority over state-of-the-art approaches. Finally, the MXF framework demonstrated its ability to capture and preserve the underlying gland structures, particularly in the base and apex regions.
3D TRUS 图像的前列腺分割是许多诊断和治疗应用的前提。不幸的是，即使对于有经验的泌尿科医生/放射科医生来说，这项艰巨的任务也存在很高的观察者内和观察者间变异性。这就是为什么自动分割算法具有重要的临床附加值。
本文介绍了一种新的深度分割架构，该架构由两个主要阶段组成：2D 切片的特定视图分割及其融合。分割阶段基于三个在特定切片观察方向上并行训练的分割网络：轴向、冠状和矢状。然后，将所提出的融合网络馈送到分割网络的输出，并对其进行训练以生成三个置信图。这些图对应于融合网络对每个特定视图分割网络的局部信任。最后，对于给定的切片，通过将这些置信图与其对应的分割组合来计算分割。通过将所有分割的切片重新堆叠以形成体积来获得前列腺的 3D 分割。
该方法在 100 名患者的数据库上进行了评估，其中包括网络架构的几种组合（分割阶段和融合阶段），以展示框架的灵活性和可靠性。所提出的方法还与 STAPLE、多数投票策略以及在同一数据库上测试的直接 3D 方法进行了比较。新方法在所有评估标准上均优于这三种方法。最后，多专家融合（MXF）框架的结果与其他最先进的方法相比具有优势，而这些方法通常在较小的数据库上工作。
我们提出了一种新的 MXF 框架来分割前列腺的 3D TRUS 图像。该方法的主要特点是在像素级别使用计算得到的置信图融合专家网络的结果。在临床数据库上进行的实验表明了该方法的稳健性和灵活性，以及其优于最先进方法的优势。最后，MXF 框架证明了其捕获和保留潜在腺体结构的能力，特别是在基底和尖端区域。
