Segmentation of lumen and vessel contours in intravascular ultrasound (IVUS) pullbacks is an arduous and time-consuming task, which demands adequately trained human resources. In the present study, we propose a machine learning approach to automatically extract lumen and vessel boundaries from IVUS datasets. The proposed approach relies on the concatenation of a deep neural network to deliver a preliminary segmentation, followed by a Gaussian process (GP) regressor to construct the final lumen and vessel contours. A multi-frame convolutional neural network (MFCNN) exploits adjacency information present in longitudinally neighboring IVUS frames, while the GP regression method filters high-dimensional noise, delivering a consistent representation of the contours. Overall, 160 IVUS pullbacks (63 patients) from the IBIS-4 study (Integrated Biomarkers and Imaging Study-4, Trial NCT00962416), were used in the present work. The MFCNN algorithm was trained with 100 IVUS pullbacks (8427 manually segmented frames), was validated with 30 IVUS pullbacks (2583 manually segmented frames) and was blindly tested with 30 IVUS pullbacks (2425 manually segmented frames). Image and contour metrics were used to characterize model performance by comparing ground truth (GT) and machine learning (ML) contours. Median values (interquartile range, IQR) of the Jaccard index for lumen and vessel were 0.913, [0.882,0.935] and 0.940, [0.917,0.957], respectively. Median values (IQR) of the Hausdorff distance for lumen and vessel were $0.196\text{mm}$, $[0.146,0.275]\text{mm}$ and $0.163\text{mm}$, $[0.122,0.234]\text{mm}$, respectively. Also, the mean value of lumen area predictions, and limits of agreement were $-0.19{\text{mm}}^2$, $[1.1,-1.5]{\text{mm}}^2$, while the mean value and limits of agreement of plaque burden were 0.0022, $[0.082,-0.078]$. The results obtained with the model developed in this work allow us to conclude that the proposed machine learning approach delivers accurate segmentations in terms of image metrics, contour metrics and clinically relevant variables, enabling its use in clinical routine by mitigating the costs involved in the manual management of IVUS datasets.

在血管内超声 (IVUS) 拉回中分割管腔和血管轮廓是一项艰巨且耗时的任务，需要经过充分培训的人力资源。在本研究中，我们提出了一种机器学习方法，可以从 IVUS 数据集中自动提取管腔和血管边界。所提出的方法依赖于深度神经网络的连接来提供初步分割，然后是高斯过程 (GP) 回归器来构建最终的管腔和血管轮廓。多帧卷积神经网络 (MFCNN) 利用纵向相邻 IVUS 帧中存在的邻接信息，而 GP 回归方法过滤高维噪声，提供轮廓的一致表示。全面的，IBIS-4 研究（Integrated Biomarkers and Imaging Study-4, Trial NCT00962416）中的 160 次 IVUS 回撤（63 名患者）用于本工作。MFCNN 算法使用 100 个 IVUS 回调（8427 个手动分割帧）进行训练，使用 30 个 IVUS 回调（2583 个手动分割帧）进行验证，并使用 30 个 IVUS 回调（2425 个手动分割帧）进行盲测。图像和轮廓度量用于通过比较地面实况 (GT) 和机器学习 (ML) 轮廓来表征模型性能。管腔和血管的 Jaccard 指数的中值（四分位距，IQR）分别为 0.913，[0.882,0.935] 和 0.940，[0.917,0.957]。管腔和血管的 Hausdorff 距离的中值 (IQR) 为 MFCNN 算法使用 100 个 IVUS 回调（8427 个手动分割帧）进行训练，使用 30 个 IVUS 回调（2583 个手动分割帧）进行验证，并使用 30 个 IVUS 回调（2425 个手动分割帧）进行盲测。图像和轮廓度量用于通过比较地面实况 (GT) 和机器学习 (ML) 轮廓来表征模型性能。管腔和血管的 Jaccard 指数的中值（四分位距，IQR）分别为 0.913，[0.882,0.935] 和 0.940，[0.917,0.957]。管腔和血管的 Hausdorff 距离的中值 (IQR) 为 MFCNN 算法使用 100 个 IVUS 回调（8427 个手动分割帧）进行训练，使用 30 个 IVUS 回调（2583 个手动分割帧）进行验证，并使用 30 个 IVUS 回调（2425 个手动分割帧）进行盲测。图像和轮廓度量用于通过比较地面实况 (GT) 和机器学习 (ML) 轮廓来表征模型性能。管腔和血管的 Jaccard 指数的中值（四分位距，IQR）分别为 0.913，[0.882,0.935] 和 0.940，[0.917,0.957]。管腔和血管的 Hausdorff 距离的中值 (IQR) 为 图像和轮廓度量用于通过比较地面实况 (GT) 和机器学习 (ML) 轮廓来表征模型性能。管腔和血管的 Jaccard 指数的中值（四分位距，IQR）分别为 0.913，[0.882,0.935] 和 0.940，[0.917,0.957]。管腔和血管的 Hausdorff 距离的中值 (IQR) 为 图像和轮廓度量用于通过比较地面实况 (GT) 和机器学习 (ML) 轮廓来表征模型性能。管腔和血管的 Jaccard 指数的中值（四分位距，IQR）分别为 0.913，[0.882,0.935] 和 0.940，[0.917,0.957]。管腔和血管的 Hausdorff 距离的中值 (IQR) 为$0.196\text{毫米}$,$[0.146,0.275]\text{毫米}$和$0.163\text{毫米}$,$[0.122,0.234]\text{毫米}$， 分别。此外，流明面积预测的平均值和一致性限制为$-0.19{\text{毫米}}^2$,$[1.1,-1.5]{\text{毫米}}^2$，而斑块负荷的平均值和一致性极限为 0.0022，$[0.082,-0.078]$. 通过这项工作中开发的模型获得的结果使我们能够得出结论，所提出的机器学习方法在图像度量、轮廓度量和临床相关变量方面提供了准确的分割，通过降低手册中涉及的成本使其能够在临床常规中使用IVUS 数据集的管理。