Objective:To explore the feasibility and accuracy of three-dimensional (3D) modeling methods based on ultrasound imaging data for normal and abnormal fetal cardiac structures, and to construct a methodology system for 3D printing of fetal heart based on ultrasound.Methods:A total of 93 fetuses examined in Tangdu Hospital of Air Force Military Medical University from January to December 2019 were selected. Fetal echocardiography was obtained using spatio-temporal image correlation (STIC). Ninety-three hearts were 3D modeled by blood flow modeling, blood pool modeling and cavity modeling, and printed by stereolithography technique. The data measured on the 3D digital models and 3D printed solid models were compared with the corresponding fetal echocardiographic images respectively in order to evaluate the accuracy of the modeling methods.Results:The fetal cardiac blood flow models based on Doppler flow image data showed the malformation and trend of small blood vessels. The fetal cardiac structure models printed based on blood pool modeling displayed the malformation of heart and large blood vessels. Models printed based on cavity modeling method accurately displayed valve and structural defects.For 83 normal fetal hearts, the long diameters of left and right ventricles measured on echocardiography [(15.3±1.9)mm, (13.2±1.9)mm] were compared with those measured on digital models [(15.1±1.9)mm, (12.9±1.9)mm] and 3D printed models[(15.1±1.9)mm, (13.0±1.9)mm], respectively, and there were no significant differences between any two groups of them ( P>0.05). Bland-Altman showed good consistency for all measurements within and between operators. Conclusions:The three modeling methods, including blood flow modeling, blood pool modeling and cavity modeling, have their own advantages in displaying different types of fetal heart malformations. Appropriate modeling methods should be selected for 3D modeling and printing to make up for the limitations of single modeling method. The consistency between measurements on 3D models and those on echocardiography is high, and the repeatability between operators is good.

Objective    To explore the safety and effectiveness of a precise marking method based on body surface mesh and three-dimensional (3D) image reconstruction. Methods    We retrospectively analyzed the clinical data of 22 patients in our hospital from October 2018 to October 2019. There were 13 males and 9 females aged 58.5 (37-72) years. All patients underwent a precise marking of pulmonary nodules based on body surface mesh and 3D image reconstruction. Then, video-assisted thoracoscopic surgery (VATS) was performed to resect the nodules. The clinical data, including positioning success rate and operation time were analyzed. Results    A total of 22 small pulmonary nodules were removed. The average diameter of small nodules was 12±3 mm, and the average distance from the visceral pleura was 17±6 mm. The localization success rate was 86.4%. The operation time was 110±43 min, and there was no surgery-related complication. Conclusion    The method of marking pulmonary nodules based on body surface mesh and 3D image reconstruction is a safe and reliable technology, which reduces the risk of hemopneumothorax caused by CT-guided lung puncture.