The Department of Cardiology, The Second Affiliated Hospital of Zhejiang University School of Medicine, Hangzhou, China.
ArteryFlow Technology Co., Ltd, Hangzhou, China.
Biomed Eng Online. 2023 Nov 28;22(1):111. doi: 10.1186/s12938-023-01171-2.
The aim of this study was to investigate the generalization performance of deep learning segmentation models on a large cohort intravascular ultrasound (IVUS) image dataset over the lumen and external elastic membrane (EEM), and to assess the consistency and accuracy of automated IVUS quantitative measurement parameters.
A total of 11,070 IVUS images from 113 patients and pullbacks were collected and annotated by cardiologists to train and test deep learning segmentation models. A comparison of five state of the art medical image segmentation models was performed by evaluating the segmentation of the lumen and EEM. Dice similarity coefficient (DSC), intersection over union (IoU) and Hausdorff distance (HD) were calculated for the overall and for subsets of different IVUS image categories. Further, the agreement between the IVUS quantitative measurement parameters calculated by automatic segmentation and those calculated by manual segmentation was evaluated. Finally, the segmentation performance of our model was also compared with previous studies.
CENet achieved the best performance in DSC (0.958 for lumen, 0.921 for EEM) and IoU (0.975 for lumen, 0.951 for EEM) among all models, while Res-UNet was the best performer in HD (0.219 for lumen, 0.178 for EEM). The mean intraclass correlation coefficient (ICC) and Bland-Altman plot demonstrated the extremely strong agreement (0.855, 95% CI 0.822-0.887) between model's automatic prediction and manual measurements.
Deep learning models based on large cohort image datasets were capable of achieving state of the art (SOTA) results in lumen and EEM segmentation. It can be used for IVUS clinical evaluation and achieve excellent agreement with clinicians on quantitative parameter measurements.
本研究旨在探讨深度学习分割模型在大样本血管内超声（IVUS）图像数据集上对管腔和外弹性膜（EEM）的泛化性能，并评估自动 IVUS 定量测量参数的一致性和准确性。
共收集了 113 名患者的 11070 张 IVUS 图像和拉回图像，并由心脏病专家进行注释，以训练和测试深度学习分割模型。通过评估管腔和 EEM 的分割，对五种最先进的医学图像分割模型进行了比较。计算了整体和不同 IVUS 图像类别子集的 Dice 相似系数（DSC）、交并比（IoU）和 Hausdorff 距离（HD）。进一步评估了自动分割计算的 IVUS 定量测量参数与手动分割计算的参数之间的一致性。最后，还将我们的模型的分割性能与之前的研究进行了比较。
CENet 在 DSC（管腔 0.958，EEM 0.921）和 IoU（管腔 0.975，EEM 0.951）方面表现最好，而 Res-UNet 在 HD（管腔 0.219，EEM 0.178）方面表现最好。组内相关系数（ICC）均值和 Bland-Altman 图表明，模型的自动预测与手动测量之间具有极强的一致性（0.855，95%置信区间 0.822-0.887）。
基于大样本图像数据集的深度学习模型能够在管腔和 EEM 分割方面达到最新技术水平（SOTA）。它可以用于 IVUS 临床评估，并在定量参数测量方面与临床医生达成极好的一致性。
