ABSTRACT
Echocardiography (echo) is gaining popularity to guide the catheter during surgical procedures. However, it is difficult to discern the catheter tip in echo even with an acoustically active catheter. An acoustically active catheter is detected for the first time in cardiac echo images using two methods. First, a convolutional neural network (CNN) model was trained to detect the region of interest (ROI), the interior of the left ventricle, containing the catheter tip. Color intensity difference detection technique was implemented on the ROI to detect the catheter. This method succeeded in detecting the catheter without any manual input on 94% and 57% of long- and short-axis projections, respectively. Second, several tracking methods were implemented and tested. Given the manually identified initial positions of the catheter, the tracking methods could distinguish between the target (catheter tip) and the surrounding on the rest of the frames. Combining the two techniques, for the first time, resulted in an automatic, robust, and fast method for catheter detection in echo images.
Subject(s)
Algorithms , Neural Networks, Computer , Catheters , Echocardiography , HeartABSTRACT
Percutaneous transendocardial injections of therapeutic agents into the myocardium may not always be effective. We used an animal model for assessing the efficacy of the injections using linoleic acid as a testing agent. Efficacious delivery into the myocardium of a beating heart was indicated by rapidly developed local myocardial necrosis and wall motion abnormalities using echocardiography. We employed this experimental model to test our innovative technology, an acoustically active injection catheter. The Doppler ultrasound-guided acoustically active injection catheter effectively delivers the substance to the myocardium but needs further technical improvements to minimize an unwanted systemic distribution of the agent.
Subject(s)
Cardiac Catheterization , Catheters , Animals , Disease Models, Animal , Injections , Ultrasonography, DopplerABSTRACT
BACKGROUND: Two-dimensional echocardiography (2D echo) is the most widely used non-invasive imaging modality due to its fast acquisition time, low cost, and high temporal resolution. Boundary identification of left ventricle (LV) in 2D echo, i.e., image segmentation, is the first step to calculate relevant clinical parameters. Currently, LV segmentation in 2D echo is primarily conducted semi-manually. A fully-automatic segmentation of the LV wall needs further development. METHODS: We evaluated the performance of the state-of-the-art convolutional neural networks (CNNs) for the segmentation of 2D echo images from 6 standard projections of the LV. We used two segmentation algorithms: U-net and segAN. The models were trained using an in-house dataset, which consists of 1,649 porcine images from 6 to 8 different pigs. In addition, a transfer learning approach was used for the segmentation of long-axis projections by training models with our database based on the previously trained weights obtained from Cardiac Acquisitions for Multi-structure Ultrasound Segmentation (CAMUS) dataset. The models were tested on a separate set of images from two other pigs by computing several metrics. The segmentation process was combined with a 3D reconstruction framework to quantify the physiological indices such as LV volumes and ejection fraction (EF). RESULTS: The average dice metric for the LV cavity was 0.90 and 0.91 for the U-net and segAN, respectively, which was higher than 0.82 for the level-set (P value: 3.31×10-25). The average Hausdorff distance for the LV cavity was 2.71 mm and 2.82 mm for the U-net and segAN, respectively, which was lower than 3.64 mm for the level-set (P value: 4.86×10-16). The LV shapes and volumes obtained using the CNN segmentation models were in good agreement with the results segmented by the experts. In addition, the differences of the calculated physiological parameters between two 3D reconstruction models segmented by the experts and CNNs were less than 15%. CONCLUSIONS: The results showed that both CNN models achieve higher performance on LV segmentation than the level-set method. The error of the reconstruction from automatic segmentation compared to the expert segmentation is less than 15%, which is within the 20% error of echo compared to the gold standard.
ABSTRACT
B-mode ultrasound imaging guidance of cannulas can be compromised by noise, artifacts, and echogenicity that is not distinctive from that of surrounding anatomy. We have modified a venovenous extracorporeal membrane oxygenation cannula by embedding piezoelectric crystals into each of its 3 blood flow ports. Each vibrating crystal acoustically interacts with a Doppler imaging signal and produces an instantaneous color marker. The aim of this study was to compare identification of the extracorporeal membrane oxygenation cannula ports by B-mode imaging versus the color Doppler marker. Unlike B-mode imaging, the color Doppler marker identified the corresponding port even in highly challenging closed-chest scans in anesthetized pigs. The method could improve guidance accuracy of cannulas by ultrasound scans.
