Segmentation of common and internal carotid arteries from 3D ultrasound images based on adaptive triple loss

Jiang, Mingjie; Zhao, Yuan; Chiu, Bernard

doi:10.1002/mp.15127

Cited by 16 publications

(9 citation statements)

References 54 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Stroke is the leading cause of morbidity and mortality throughout the world. 1 Over 80% of strokes are ischemic, many resulting from the blockage of a cerebral artery by atherosclerotic emboli. Carotid atherosclerosis is a major source of emboli, generated from plaque rupture as platelet aggregates or plaque debris.…”

Section: Purposementioning

confidence: 99%

A freehand 3D ultrasound carotid scanning system for point-of-care ultrasonography

Ren

Chiu

2023

Medical Imaging 2023: Ultrasonic Imaging and Tomography

View full text Add to dashboard Cite

Stroke is a leading cause of morbidity and mortality throughout the world. Three-dimensional ultrasound (3DUS) imaging was shown to be more sensitive to treatment effect and more accurate in stratifying stroke risk than two-dimensional ultrasound (2DUS) imaging. Point-of-care ultrasound screening (POCUS) is important for patients with limited mobility and at times when the patients have limited access to the ultrasound scanning room, such as in the COVID-19 era. We used an optical tracking system to track the 3D position and orientation of the 2DUS frames acquired by a commercial wireless ultrasound system and subsequently reconstructed a 3DUS image from these frames. The tracking requires spatial and temporal calibrations. Spatial calibration is required to determine the spatial relationship between the 2DUS machine and the tracking system. Spatial calibration was achieved by localizing the landmarks with known coordinates in a custom-designed Z-fiducial phantom in an 2DUS image. Temporal calibration is needed to synchronize the clock of the wireless ultrasound system and the optical tracking system so that position and orientation detected by the optical tracking system can be registered to the corresponding 2DUS frame. Temporal calibration was achieved by initiating the scanning by an abrupt motion that can be readily detected in both systems. This abrupt motion establishes a common reference time point, thereby synchronizing the clock in both systems. We demonstrated that the system can be used to visualize the three-dimensional structure of a carotid phantom. The error rate of the measurements is 2.3%. Upon in-vivo validation, this system will allow POCUS carotid scanning in clinical research and practices.

show abstract

Section: Purposementioning

confidence: 99%

A freehand 3D ultrasound carotid scanning system for point-of-care ultrasonography

Ren

Chiu

2023

Medical Imaging 2023: Ultrasonic Imaging and Tomography

View full text Add to dashboard Cite

show abstract

“…The automatic segmentation of the 3D carotid US images was performed with a twochannel U-Net, driven by the ATDL function proposed previously [33]. The structure of the network is shown in Figure 3.…”

Section: Segmentation Frameworkmentioning

confidence: 99%

“…The two-channel network can segment LIB and MAB simultaneously, which is more efficient than using two independent models for MAB and LIB segmentations. A single network with two channels is more able to model the geometric relationship between the MAB and LIB (e.g., the LIB is always inside the MAB) than two independent networks segmenting the MAB and LIB separately [33]. The convolutional module consists of two convolutional layers with kernel size 3 × 3 and stride 1, each followed by a batch normalization (BN) layer and a ReLu activation layer.…”

Section: Segmentation Frameworkmentioning

confidence: 99%

“…Due to the overlap between MAB and LIB, we use the sigmoid function instead of softmax as the activation function of the last layer because softmax requires the sum of the probabilities in all channels to be 1. The ATDL used for supervising the training of the U-net has been described elsewhere [33], and briefly summarized here. ATDL is the adaptive Dice loss associated with the three components: MAB, LIB, and carotid vessel wall (CVW):…”

Section: Segmentation Frameworkmentioning

confidence: 99%

“…VWV and VWT measured only at CCA may be less sensitive to treatment effect than those measured in previous studies that involved both CCA and ICA [16,22], with carotid plaques typically more prevalent at ICA [32]. Jiang et al [33] previously developed a twochannel U-Net, driven by a novel loss function called the adaptive triple Dice loss (ATDL) function that can segment LIB and MAB simultaneously. The segmentation performed on CCA slices was fully automatic, whereas the ICA was enclosed by a manually identified bounding box that ensured the ICA was segmented, instead of the external carotid artery (ECA), which are less relevant to cardiovascular events [34] and typically not measured for atherosclerosis [9,22,35].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Development of a Three-Dimensional Carotid Ultrasound Image Segmentation Workflow for Improved Efficiency, Reproducibility and Accuracy in Measuring Vessel Wall and Plaque Volume and Thickness

Zhao,

Jiang,

Chan

et al. 2023

Bioengineering

Self Cite

View full text Add to dashboard Cite

Automated segmentation of carotid lumen-intima boundary (LIB) and media-adventitia boundary (MAB) by deep convolutional neural networks (CNN) from three-dimensional ultrasound (3DUS) images has made assessment and monitoring of carotid atherosclerosis more efficient than manual segmentation. However, training of CNN still requires manual segmentation of LIB and MAB. Therefore, there is a need to improve the efficiency of manual segmentation and develop strategies to improve segmentation accuracy by the CNN for serial monitoring of carotid atherosclerosis. One strategy to reduce segmentation time is to increase the interslice distance (ISD) between segmented axial slices of a 3DUS image while maintaining the segmentation reliability. We, for the first time, investigated the effect of ISD on the reproducibility of MAB and LIB segmentations. The intra-observer reproducibility of LIB and MAB segmentations at ISDs of 1 mm and 2 mm was not statistically significantly different, whereas the reproducibility at ISD = 3 mm was statistically lower. Therefore, we conclude that segmentation with an ISD of 2 mm provides sufficient reliability for CNN training. We further proposed training the CNN by the baseline images of the entire cohort of patients for automatic segmentation of the follow-up images acquired for the same cohort. We validated that segmentation with this time-based partitioning approach is more accurate than that produced by patient-based partitioning, especially at the carotid bifurcation. This study forms the basis for an efficient, reproducible, and accurate 3DUS workflow for serial monitoring of carotid atherosclerosis useful in risk stratification of cardiovascular events and in evaluating the efficacy of new treatments.

show abstract

Unsupervised shape‐and‐texture‐based generative adversarial tuning of pre‐trained networks for carotid segmentation from 3D ultrasound images

Chen,

Jiang,

Chiu

2024

Medical Physics

View full text Add to dashboard Cite

BackgroundVessel‐wall volume and localized three‐dimensional ultrasound (3DUS) metrics are sensitive to the change of carotid atherosclerosis in response to medical/dietary interventions. Manual segmentation of the media‐adventitia boundary (MAB) and lumen‐intima boundary (LIB) required to obtain these metrics is time‐consuming and prone to observer variability. Although supervised deep‐learning segmentation models have been proposed, training of these models requires a sizeable manually segmented training set, making larger clinical studies prohibitive.PurposeWe aim to develop a method to optimize pre‐trained segmentation models without requiring manual segmentation to supervise the fine‐tuning process.MethodsWe developed an adversarial framework called the unsupervised shape‐and‐texture generative adversarial network (USTGAN) to fine‐tune a convolutional neural network (CNN) pre‐trained on a source dataset for accurate segmentation of a target dataset. The network integrates a novel texture‐based discriminator with a shape‐based discriminator, which together provide feedback for the CNN to segment the target images in a similar way as the source images. The texture‐based discriminator increases the accuracy of the CNN in locating the artery, thereby lowering the number of failed segmentations. Failed segmentation was further reduced by a self‐checking mechanism to flag longitudinal discontinuity of the artery and by self‐correction strategies involving surface interpolation followed by a case‐specific tuning of the CNN. The U‐Net was pre‐trained by the source dataset involving 224 3DUS volumes with 136, 44, and 44 volumes in the training, validation and testing sets. The training of USTGAN involved the same training group of 136 volumes in the source dataset and 533 volumes in the target dataset. No segmented boundaries for the target cohort were available for training USTGAN. The validation and testing of USTGAN involved 118 and 104 volumes from the target cohort, respectively. The segmentation accuracy was quantified by Dice Similarity Coefficient (DSC), and incorrect localization rate (ILR). Tukey's Honestly Significant Difference multiple comparison test was employed to quantify the difference of DSCs between models and settings, where was considered statistically significant.ResultsUSTGAN attained a DSC of % in LIB and % in MAB, improving from the baseline performance of % in LIB (p ) and % in MAB (p ). Our approach outperformed six state‐of‐the‐art domain‐adaptation models (MAB: , LIB: ). The proposed USTGAN also had the lowest ILR among the methods compared (LIB: 2.5%, MAB: 1.7%).ConclusionOur framework improves segmentation generalizability, thereby facilitating efficient carotid disease monitoring in multicenter trials and in clinics with less expertise in 3DUS imaging.

show abstract

Segmentation of common and internal carotid arteries from 3D ultrasound images based on adaptive triple loss

Cited by 16 publications

References 54 publications

A freehand 3D ultrasound carotid scanning system for point-of-care ultrasonography

A freehand 3D ultrasound carotid scanning system for point-of-care ultrasonography

Development of a Three-Dimensional Carotid Ultrasound Image Segmentation Workflow for Improved Efficiency, Reproducibility and Accuracy in Measuring Vessel Wall and Plaque Volume and Thickness

Unsupervised shape‐and‐texture‐based generative adversarial tuning of pre‐trained networks for carotid segmentation from 3D ultrasound images

Contact Info

Product

Resources

About