Hao Xu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Hao Xu has a total of 9 co-authors. Here are the top co-authors by number of shared publications:
- Aiyun Sun
CT Imaging Research Center, GE HealthCare China, 200010 Shanghai, China
2 shared publications - Jiangfeng Pan
Department of Medical Imaging, Affiliated Jinhua Hospital, Zhejiang University School of Medicine, 321000 Jinhua, Zhejiang, China
2 shared publications - Jing Yuan
College of Mathematical Medicine, Zhejiang Normal University, 321004 Jinhua, Zhejiang, China
2 shared publications - Ronghua Wu
Department of Medical Imaging, Affiliated Jinhua Hospital, Zhejiang University School of Medicine, 321000 Jinhua, Zhejiang, China
2 shared publications - Shanshan Ying
Department of Medical Imaging, Affiliated Jinhua Hospital, Zhejiang University School of Medicine, 321000 Jinhua, Zhejiang, China
2 shared publications - Xiaorong Chen
Department of Medical Imaging, Affiliated Jinhua Hospital, Zhejiang University School of Medicine, 321000 Jinhua, Zhejiang, China
2 shared publications
Latest Publications
- Open Access
- 18 May 2026
- Published by IMR Press in Reviews in Cardiovascular Medicine
- Vol. 27 (5)
Cardiac Structural and Functional Evaluation Using a Heart Motion Correction Algorithm for Coronary Computed Tomography Angiography in Patients With High Heart Rates
Background: The heart motion correction algorithm used in current multi-slice computed tomography (CT) is sufficient for coronary artery imaging in patients with high heart rates. However, the effect of this algorithm on the image quality in whole-cardiac-cycle reconstructions remains unclear. Therefore, this study aimed to investigate image quality, segmentation performance, and cardiac structure and function assessment using a heart motion correction algorithm for coronary CT angiography in patients with rapid heart rates. Methods: This study retrospectively collected data from 58 consecutive patients with high heart rates (≥80 beats/min), of whom 36 also underwent cardiac magnetic resonance (CMR) imaging. CT images were reconstructed from 0% to 100% in 5% increments using the standard reconstruction (STD) and second-generation snapshot freeze (SSF2) protocols, and then processed by an automatic heart segmentation algorithm. Image quality, segmentation performance, cardiac volumes, and functional parameters were compared between protocols. Results: Compared with the STD protocol, the SSF2 protocol yielded a higher image quality score (3.91 ± 0.29 vs. 3.84 ± 0.37; p < 0.01), a steeper edge rise slope (41.71 ± 19.03 vs. 25.59 ± 13.16; p < 0.01), and lower entropy (4.12 ± 0.48 vs. 4.40 ± 0.28; p < 0.01). For left ventricular end-diastolic volume, the intraclass correlation coefficient (ICC) between automatic segmentation and manual contouring for the SSF2 protocol was 0.96, and the coefficient of variation was 7.84%. In contrast, the coefficients of variation for left ventricular end-systolic volume were poor (48.24% for STD and 48.18% for SSF2). Differences in global circumferential strain (–13.30 ± 3.42 vs. –15.01 ± 4.44; p < 0.01) and global longitudinal strain (–11.80 ± 4.83 vs. –13.01 ± 4.36; p < 0.01) between SSF2 and CMR were statistically significant, although correlations (ICC = 0.90 and 0.85, respectively) were good. Conclusions: SSF2 significantly improves image quality, structure, and function, and enables strain assessment in whole-cardiac-cycle reconstructions in patients with high heart rates. SSF2 also demonstrates superior performance over the STD protocol for evaluating myocardial strain.
- 18 May 2026
- Published by IMR Press in Reviews in Cardiovascular Medicine
- Vol. 27 (5)
Open AccessCardiac Structural and Functional Evaluation Using a Heart Motion Correction Algorithm for Coronary Computed Tomography Angiography in Patients With High Heart Rates
Background: The heart motion correction algorithm used in current multi-slice computed tomography (CT) is sufficient for coronary artery imaging in patients with high heart rates. However, the effect of this algorithm on the image quality in whole-cardiac-cycle reconstructions remains unclear. Therefore, this study aimed to investigate image quality, segmentation performance, and cardiac structure and function assessment using a heart motion correction algorithm for coronary CT angiography in patients with rapid heart rates. Methods: This study retrospectively collected data from 58 consecutive patients with high heart rates (≥80 beats/min), of whom 36 also underwent cardiac magnetic resonance (CMR) imaging. CT images were reconstructed from 0% to 100% in 5% increments using the standard reconstruction (STD) and second-generation snapshot freeze (SSF2) protocols, and then processed by an automatic heart segmentation algorithm. Image quality, segmentation performance, cardiac volumes, and functional parameters were compared between protocols. Results: Compared with the STD protocol, the SSF2 protocol yielded a higher image quality score (3.91 ± 0.29 vs. 3.84 ± 0.37; p < 0.01), a steeper edge rise slope (41.71 ± 19.03 vs. 25.59 ± 13.16; p < 0.01), and lower entropy (4.12 ± 0.48 vs. 4.40 ± 0.28; p < 0.01). For left ventricular end-diastolic volume, the intraclass correlation coefficient (ICC) between automatic segmentation and manual contouring for the SSF2 protocol was 0.96, and the coefficient of variation was 7.84%. In contrast, the coefficients of variation for left ventricular end-systolic volume were poor (48.24% for STD and 48.18% for SSF2). Differences in global circumferential strain (–13.30 ± 3.42 vs. –15.01 ± 4.44; p < 0.01) and global longitudinal strain (–11.80 ± 4.83 vs. –13.01 ± 4.36; p < 0.01) between SSF2 and CMR were statistically significant, although correlations (ICC = 0.90 and 0.85, respectively) were good. Conclusions: SSF2 significantly improves image quality, structure, and function, and enables strain assessment in whole-cardiac-cycle reconstructions in patients with high heart rates. SSF2 also demonstrates superior performance over the STD protocol for evaluating myocardial strain.
- Open Access
- 18 May 2026
- Vol. 27 (5)
Cardiac Structural and Functional Evaluation Using a Heart Motion Correction Algorithm for Coronary Computed Tomography Angiography in Patients With High Heart Rates
The heart motion correction algorithm used in current multi-slice computed tomography (CT) is sufficient for coronary artery imaging in patients with high heart rates. However, the effect of this algorithm on the image quality in whole-cardiac-cycle reconstructions remains unclear. Therefore, this study aimed to investigate image quality, segmentation performance, and cardiac structure and function assessment using a heart motion correction algorithm for coronary CT angiography in patients with rapid heart rates. This study retrospectively collected data from 58 consecutive patients with high heart rates (≥80 beats/min), of whom 36 also underwent cardiac magnetic resonance (CMR) imaging. CT images were reconstructed from 0% to 100% in 5% increments using the standard reconstruction (STD) and second-generation snapshot freeze (SSF2) protocols, and then processed by an automatic heart segmentation algorithm. Image quality, segmentation performance, cardiac volumes, and functional parameters were compared between protocols. Compared with the STD protocol, the SSF2 protocol yielded a higher image quality score (3.91 ± 0.29 vs. 3.84 ± 0.37; p < 0.01), a steeper edge rise slope (41.71 ± 19.03 vs. 25.59 ± 13.16; p < 0.01), and lower entropy (4.12 ± 0.48 vs. 4.40 ± 0.28; p < 0.01). For left ventricular end-diastolic volume, the intraclass correlation coefficient (ICC) between automatic segmentation and manual contouring for the SSF2 protocol was 0.96, and the coefficient of variation was 7.84%. In contrast, the coefficients of variation for left ventricular end-systolic volume were poor (48.24% for STD and 48.18% for SSF2). Differences in global circumferential strain (–13.30 ± 3.42 vs. –15.01 ± 4.44; p < 0.01) and global longitudinal strain (–11.80 ± 4.83 vs. –13.01 ± 4.36; p < 0.01) between SSF2 and CMR were statistically significant, although correlations (ICC = 0.90 and 0.85, respectively) were good. SSF2 significantly improves image quality, structure, and function, and enables strain assessment in whole-cardiac-cycle reconstructions in patients with high heart rates. SSF2 also demonstrates superior performance over the STD protocol for evaluating myocardial strain.
- 18 May 2026
- Vol. 27 (5)
Open AccessCardiac Structural and Functional Evaluation Using a Heart Motion Correction Algorithm for Coronary Computed Tomography Angiography in Patients With High Heart Rates
The heart motion correction algorithm used in current multi-slice computed tomography (CT) is sufficient for coronary artery imaging in patients with high heart rates. However, the effect of this algorithm on the image quality in whole-cardiac-cycle reconstructions remains unclear. Therefore, this study aimed to investigate image quality, segmentation performance, and cardiac structure and function assessment using a heart motion correction algorithm for coronary CT angiography in patients with rapid heart rates. This study retrospectively collected data from 58 consecutive patients with high heart rates (≥80 beats/min), of whom 36 also underwent cardiac magnetic resonance (CMR) imaging. CT images were reconstructed from 0% to 100% in 5% increments using the standard reconstruction (STD) and second-generation snapshot freeze (SSF2) protocols, and then processed by an automatic heart segmentation algorithm. Image quality, segmentation performance, cardiac volumes, and functional parameters were compared between protocols. Compared with the STD protocol, the SSF2 protocol yielded a higher image quality score (3.91 ± 0.29 vs. 3.84 ± 0.37; p < 0.01), a steeper edge rise slope (41.71 ± 19.03 vs. 25.59 ± 13.16; p < 0.01), and lower entropy (4.12 ± 0.48 vs. 4.40 ± 0.28; p < 0.01). For left ventricular end-diastolic volume, the intraclass correlation coefficient (ICC) between automatic segmentation and manual contouring for the SSF2 protocol was 0.96, and the coefficient of variation was 7.84%. In contrast, the coefficients of variation for left ventricular end-systolic volume were poor (48.24% for STD and 48.18% for SSF2). Differences in global circumferential strain (–13.30 ± 3.42 vs. –15.01 ± 4.44; p < 0.01) and global longitudinal strain (–11.80 ± 4.83 vs. –13.01 ± 4.36; p < 0.01) between SSF2 and CMR were statistically significant, although correlations (ICC = 0.90 and 0.85, respectively) were good. SSF2 significantly improves image quality, structure, and function, and enables strain assessment in whole-cardiac-cycle reconstructions in patients with high heart rates. SSF2 also demonstrates superior performance over the STD protocol for evaluating myocardial strain.