Methods of heart data based on ultrasound-CT cardiac image fusion
10.13929/j.1003-3289.201901202
- Author:
Sikai CHEN
1
Author Information
1. Department of Ultrasound Imaging, Renmin Hospital, Wuhan University
- Publication Type:Journal Article
- Keywords:
Animal experimentation;
Heart;
Image fusion;
Tomography, X-ray computed;
Ultrasonography
- From:
Chinese Journal of Medical Imaging Technology
2019;35(9):1310-1314
- CountryChina
- Language:Chinese
-
Abstract:
Objective: To obtain ultrasound-CT cardiac imaging fusion methods and test indicators by using multi-modal cardiac image fusion technology during animal experiment, and to verify the feasibility of this method with clinical experiment. Methods: For animal experiment, heart CT images and ultrasound heart valve volume images of 6 Beagle dogs were collected. The alignment points and the annulus planes were aligned using Mimics innovationsuite 19.0 and 3-Matic 11.0 software, and ultrasound heart valve images were fused to CT heart chamber, the registration index was measured in the fused image and the cardiac specimen. The index included the maximum diameter of anterior lateral combination (ALC) of mitral valve to apex (D-ALC), the posterior interior combination (PIC) of mitral valve to apex (D-PIC) and the aortic valve closure (AVC) to apex (D-AVC), and the angle between the plane of mitral annulus and aortic annulus (AMA). In clinical trials, 41 patients with atrial fibrillation underwent CT and transesophageal echocardiography. The images were processed in the same way, and the registration indexes were measured based on CT images and CT-ultrasound fusion images. Results: In animal experiment, the ultrasound-CT images were successfully registered. There was no significant difference in D-ALC, D-PIC, D-AVC nor AMA between the fusion image and the cardiac specimen (all P>0.05). In clinical trials, ultrasound-CT cardiac image fusion registration was successfully performed in all 41 patients. There was no significant difference of D-ALC, D-PIC, D-AVC nor AMA between clinical fusion images and clinical CT images (all P>0.05). After image fusion, 92.68% (38/41) of AMA difference was within 5°, and 87.80% (36/41) of D-ALC, D-PIC, D-AVC difference percentages were within 5%. Conclusion: Cardiac CT-ultrasound image fusion is feasible and accurate using the internal feature of the annulus as an anatomical landmark. The AMA and the distance between the annulus plane and the apex can be used as detection indexes for registration accuracy.
