Hongbo Zhang
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Hongbo Zhang has a total of 28 co-authors. Here are the top co-authors by number of shared publications:
- Xiaohai Ma
Department of Interventional Diagnosis and Treatment, Beijing Anzhen Hospital, Capital Medical University, Beijing, China
4 shared publications - Lei Zhao
Department of Radiology, Beijing Anzhen Hospital, Capital Medical University, Beijing, China
3 shared publications Chen Zhang
Department of Interventional Diagnosis and Treatment Beijing Anzhen Hospital Affiliated to Capital Medical University Beijing China
2 shared publications- Chen Zhang1 shared publication
- Gang Deng
Department of Radiology
1 shared publication - Guanyu Lu
Department of Interventional Diagnosis and Treatment Beijing Anzhen Hospital Affiliated to Capital Medical University Beijing China
1 shared publication
Latest Publications
- 23 April 2026
- Published by Wiley in Journal of Magnetic Resonance Imaging
- Vol. 64 (1) , 153-165
Large Language Models for Cardiac MRI Diagnosis Based on Standardized Text Descriptions
Background: MRI is important for cardiac disease evaluation, but accurate diagnosis remains challenging in less experienced centers. Although large language models (LLMs) have shown promise in medical imaging diagnosis, their application in cardiac MRI is limited. Hypothesis: LLMs may be effective in achieving cardiac MRI diagnosis based on standardized descriptions. Study Type: Retrospective. Population: A total of 203 hypertrophic cardiomyopathy, 186 dilated cardiomyopathy, 46 hypertensive heart disease, 198 ischemic cardiomyopathy, 38 constrictive pericarditis, 45 cardiac amyloidosis, 91 myocarditis, and 144 normal controls. Field Strength/Sequences: Balanced steady‐state free‐precession, short tau inversion recovery, and breath‐hold inversion‐recovery segmented gradient‐echo sequences at 3.0 T. Assessment: Clinical and cardiac MRI information from each subject was converted into standardized descriptions and input into Generative Pre‐trained Transformer‐4.5 (GPT‐4.5), GPT‐4 Omni (GPT‐4o), Deepseek‐V3, and Deepseek‐R1 LLMs. Cardiac MRI information included LV function, wall thickness and motion, and abnormalities on T2WI, perfusion and late gadolinium enhancement sequences. Each model was asked to generate an imaging diagnosis. In addition, a medical student (8 months experience) and three radiologists (junior, mid‐level and senior: with 3, 6, and 10 years' experience, respectively) provided diagnoses based on cardiac MRI images and clinical information. Statistic Tests: Frequency‐weighted sensitivity and specificity were calculated. The diagnostic performances of the LLMs and human readers were compared using the McNemar test with Bonferroni correction. A p value < 0.05 was considered significant. Results: All LLMs showed excellent frequency‐weighted specificity (0.973–0.983). The frequency‐weighted sensitivities of all LLMs were not significantly different from that of the junior radiologist, were significantly higher than that of the medical student, and significantly inferior to those of the senior radiologist (GPT‐4.5: 0.863, GPT‐4o: 0.821, Deepseek‐V3: 0.843, and Deepseek‐R1: 0.851 vs. junior radiologist: 0.850, all adjusted p = 1.000; vs. medical student: 0.731, all adjusted p < 0.001; vs. senior radiologist: 0.942, all adjusted p < 0.001). Additionally, the mid‐level radiologist achieved a frequency‐weighted sensitivity of 0.895, outperforming all LLMs except GPT‐4.5. Data Conclusion: LLMs may generate accurate diagnoses from standardized cardiac MRI descriptions, potentially benefiting less experienced physicians. Level of Evidence: 4. Technical Efficacy: Stage 5.
- 23 April 2026
- Published by Wiley in Journal of Magnetic Resonance Imaging
- Vol. 64 (1) , 153-165
Large Language Models for Cardiac MRI Diagnosis Based on Standardized Text Descriptions
Background: MRI is important for cardiac disease evaluation, but accurate diagnosis remains challenging in less experienced centers. Although large language models (LLMs) have shown promise in medical imaging diagnosis, their application in cardiac MRI is limited. Hypothesis: LLMs may be effective in achieving cardiac MRI diagnosis based on standardized descriptions. Study Type: Retrospective. Population: A total of 203 hypertrophic cardiomyopathy, 186 dilated cardiomyopathy, 46 hypertensive heart disease, 198 ischemic cardiomyopathy, 38 constrictive pericarditis, 45 cardiac amyloidosis, 91 myocarditis, and 144 normal controls. Field Strength/Sequences: Balanced steady‐state free‐precession, short tau inversion recovery, and breath‐hold inversion‐recovery segmented gradient‐echo sequences at 3.0 T. Assessment: Clinical and cardiac MRI information from each subject was converted into standardized descriptions and input into Generative Pre‐trained Transformer‐4.5 (GPT‐4.5), GPT‐4 Omni (GPT‐4o), Deepseek‐V3, and Deepseek‐R1 LLMs. Cardiac MRI information included LV function, wall thickness and motion, and abnormalities on T2WI, perfusion and late gadolinium enhancement sequences. Each model was asked to generate an imaging diagnosis. In addition, a medical student (8 months experience) and three radiologists (junior, mid‐level and senior: with 3, 6, and 10 years' experience, respectively) provided diagnoses based on cardiac MRI images and clinical information. Statistic Tests: Frequency‐weighted sensitivity and specificity were calculated. The diagnostic performances of the LLMs and human readers were compared using the McNemar test with Bonferroni correction. A p value < 0.05 was considered significant. Results: All LLMs showed excellent frequency‐weighted specificity (0.973–0.983). The frequency‐weighted sensitivities of all LLMs were not significantly different from that of the junior radiologist, were significantly higher than that of the medical student, and significantly inferior to those of the senior radiologist (GPT‐4.5: 0.863, GPT‐4o: 0.821, Deepseek‐V3: 0.843, and Deepseek‐R1: 0.851 vs. junior radiologist: 0.850, all adjusted p = 1.000; vs. medical student: 0.731, all adjusted p < 0.001; vs. senior radiologist: 0.942, all adjusted p < 0.001). Additionally, the mid‐level radiologist achieved a frequency‐weighted sensitivity of 0.895, outperforming all LLMs except GPT‐4.5. Data Conclusion: LLMs may generate accurate diagnoses from standardized cardiac MRI descriptions, potentially benefiting less experienced physicians. Level of Evidence: 4. Technical Efficacy: Stage 5.
- 25 April 2024
- Published by Wiley in Journal of Magnetic Resonance Imaging
- Vol. 61 (1) , 319-320
Editorial for “Second‐Order Motion‐Compensated Echo‐Planar Cardiac Diffusion‐Weighted MRI: Usefulness of Compressed Sensitivity Encoding”
- 25 April 2024
- Published by Wiley in Journal of Magnetic Resonance Imaging
- Vol. 61 (1) , 319-320
Editorial for “Second‐Order Motion‐Compensated Echo‐Planar Cardiac Diffusion‐Weighted MRI: Usefulness of Compressed Sensitivity Encoding”
- 1 February 2024
- Published by Radiological Society of North America (RSNA) in Radiology: Cardiothoracic Imaging
- Vol. 6 (1) , e230323
Radiomics from Cardiovascular MR Cine Images for Identifying Patients with Hypertrophic Cardiomyopathy at High Risk for Heart Failure
Purpose To develop a model integrating radiomics features from cardiac MR cine images with clinical and standard cardiac MRI predictors to identify patients with hypertrophic cardiomyopathy (HCM) at high risk for heart failure (HF). Materials and Methods In this retrospective study, 516 patients with HCM (median age, 51 years [IQR: 40-62]; 367 [71.1%] men) who underwent cardiac MRI from January 2015 to June 2021 were divided into training and validation sets (7:3 ratio). Radiomics features were extracted from cardiac cine images, and radiomics scores were calculated based on reproducible features using the least absolute shrinkage and selection operator Cox regression. Radiomics scores and clinical and standard cardiac MRI predictors that were significantly associated with HF events in univariable Cox regression analysis were incorporated into a multivariable analysis to construct a combined prediction model. Model performance was validated using time-dependent area under the receiver operating characteristic curve (AUC), and the optimal cutoff value of the combined model was determined for patient risk stratification. Results The radiomics score was the strongest predictor for HF events in both univariable (hazard ratio, 10.37; P < .001) and multivariable (hazard ratio, 10.25; P < .001) analyses. The combined model yielded the highest 1- and 3-year AUCs of 0.81 and 0.80, respectively, in the training set and 0.82 and 0.77 in the validation set. Patients stratified as high risk had more than sixfold increased risk of HF events compared with patients at low risk. Conclusion The combined model with radiomics features and clinical and standard cardiac MRI parameters accurately identified patients with HCM at high risk for HF. Keywords: Cardiomyopathies, Outcomes Analysis, Cardiovascular MRI, Hypertrophic Cardiomyopathy, Radiomics, Heart Failure Supplemental material is available for this article. © RSNA, 2024.
- 1 February 2024
- Published by Radiological Society of North America (RSNA) in Radiology: Cardiothoracic Imaging
- Vol. 6 (1) , e230323
Radiomics from Cardiovascular MR Cine Images for Identifying Patients with Hypertrophic Cardiomyopathy at High Risk for Heart Failure
Purpose To develop a model integrating radiomics features from cardiac MR cine images with clinical and standard cardiac MRI predictors to identify patients with hypertrophic cardiomyopathy (HCM) at high risk for heart failure (HF). Materials and Methods In this retrospective study, 516 patients with HCM (median age, 51 years [IQR: 40-62]; 367 [71.1%] men) who underwent cardiac MRI from January 2015 to June 2021 were divided into training and validation sets (7:3 ratio). Radiomics features were extracted from cardiac cine images, and radiomics scores were calculated based on reproducible features using the least absolute shrinkage and selection operator Cox regression. Radiomics scores and clinical and standard cardiac MRI predictors that were significantly associated with HF events in univariable Cox regression analysis were incorporated into a multivariable analysis to construct a combined prediction model. Model performance was validated using time-dependent area under the receiver operating characteristic curve (AUC), and the optimal cutoff value of the combined model was determined for patient risk stratification. Results The radiomics score was the strongest predictor for HF events in both univariable (hazard ratio, 10.37; P < .001) and multivariable (hazard ratio, 10.25; P < .001) analyses. The combined model yielded the highest 1- and 3-year AUCs of 0.81 and 0.80, respectively, in the training set and 0.82 and 0.77 in the validation set. Patients stratified as high risk had more than sixfold increased risk of HF events compared with patients at low risk. Conclusion The combined model with radiomics features and clinical and standard cardiac MRI parameters accurately identified patients with HCM at high risk for HF. Keywords: Cardiomyopathies, Outcomes Analysis, Cardiovascular MRI, Hypertrophic Cardiomyopathy, Radiomics, Heart Failure Supplemental material is available for this article. © RSNA, 2024.
- Open Access
Development and external validation of a dynamic nomogram for delayed cerebral ischaemia after aneurysmal subarachnoid hemorrhage: a study protocol for a multicentre retrospective cohort study
Introduction Delayed cerebral ischaemia (DCI) caused by aneurysmal subarachnoid haemorrhage (aSAH) is the most frequent complication and typically contributes to poor neurological outcome or deterioration of patients’ condition. Therefore, early accurate and effective prediction of DCI is urgently needed. This study aims to construct a dynamic nomogram for precisely calculating the risk of DCI in patients with aSAH. Internal validation of this tool is conducted using the training cohort, and independent external validation is completed by using other medical centre datasets. Methods and analysis This study is a multicentre, retrospective, observational cohort study using data from patients with aSAH. The participants include all adult patients who received surgical treatment in neurosurgery of multiple medical centres from 1 September 2019 to 1 April 2021, including Renmin Hospital of Wuhan University, Huzhou Central Hospital, First Affiliated Hospital of Harbin Medical University, General Hospital of Northern Theatre Command and Affiliated Hospital of Panzhihua University. Clinical information is collected via the electronic medical record system, including demographic data, clinical state on admission and serum laboratory tests. Modified Fisher grade at admission, admission subarachnoid clot and cerebral oedema density, and residual postoperative subarachnoid clot density are determined using the electronic imagine record software. The primary outcome is DCI. Ethics and dissemination This study protocol was reviewed and approved by the Medical Ethics Committee of Renmin Hospital of Wuhan University, which is the principal affiliation of this study (approval number: WDRM2021-K022). The other Ethics Committees, including Huzhou Central Hospital (approval number: 202108005–01), First Affiliated Hospital of Harbin Medical University (approval number: H202156), General Hospital of Northern Theater Command (approval number: Y2021060) and Affiliated Hospital of Panzhihua University (approval number: 202105002), also approved the protocol. The results of this research will be published in a peer-reviewed medical journal. Trial registration number ChiCTR2100044448.
Open AccessDevelopment and external validation of a dynamic nomogram for delayed cerebral ischaemia after aneurysmal subarachnoid hemorrhage: a study protocol for a multicentre retrospective cohort study
Introduction Delayed cerebral ischaemia (DCI) caused by aneurysmal subarachnoid haemorrhage (aSAH) is the most frequent complication and typically contributes to poor neurological outcome or deterioration of patients’ condition. Therefore, early accurate and effective prediction of DCI is urgently needed. This study aims to construct a dynamic nomogram for precisely calculating the risk of DCI in patients with aSAH. Internal validation of this tool is conducted using the training cohort, and independent external validation is completed by using other medical centre datasets. Methods and analysis This study is a multicentre, retrospective, observational cohort study using data from patients with aSAH. The participants include all adult patients who received surgical treatment in neurosurgery of multiple medical centres from 1 September 2019 to 1 April 2021, including Renmin Hospital of Wuhan University, Huzhou Central Hospital, First Affiliated Hospital of Harbin Medical University, General Hospital of Northern Theatre Command and Affiliated Hospital of Panzhihua University. Clinical information is collected via the electronic medical record system, including demographic data, clinical state on admission and serum laboratory tests. Modified Fisher grade at admission, admission subarachnoid clot and cerebral oedema density, and residual postoperative subarachnoid clot density are determined using the electronic imagine record software. The primary outcome is DCI. Ethics and dissemination This study protocol was reviewed and approved by the Medical Ethics Committee of Renmin Hospital of Wuhan University, which is the principal affiliation of this study (approval number: WDRM2021-K022). The other Ethics Committees, including Huzhou Central Hospital (approval number: 202108005–01), First Affiliated Hospital of Harbin Medical University (approval number: H202156), General Hospital of Northern Theater Command (approval number: Y2021060) and Affiliated Hospital of Panzhihua University (approval number: 202105002), also approved the protocol. The results of this research will be published in a peer-reviewed medical journal. Trial registration number ChiCTR2100044448.
- 28 September 2021
- Published by Wiley in Journal of Magnetic Resonance Imaging
- Vol. 55 (6) , 1828-1840
Quantification of Myocardial Deformation in Patients with Takayasu Arteritis by Cardiovascular Magnetic Resonance Feature Tracking Imaging
Background: Cardiac involvement is one of the main causes of morbidity and mortality in patients with Takayasu arteritis (TA). Early detection and intervention of cardiac damage may be helpful to reduce the mortality of TA. Magnetic resonance (MR) imaging (MRI)‐derived feature tracking (FT) is an effective quantitative method to assess myocardial deformation which may reflect early changes of cardiac function.Purpose: To explore the utility of MR‐FT as a method to detect cardiac damage in TA patients.Study Type: Retrospective.Population: Fifty‐seven TA patients who had undergone clinically indicated MRI and 57 healthy controls.Field Strength/Sequences: Balanced steady‐state free precession rest cine and 2D phase‐sensitive inversion recovery breath‐hold segmented gradient echo late gadolinium enhancement (LGE) sequences at 3.0 T.Assessment: Based on LGE images, TA patients were divided into two subgroups, LGE (+) subgroup (N = 12) and LGE (−) subgroup (N = 45). In addition, patients were further subdivided into impaired (N = 26) and preserved left ventricle ejection fraction (LVEF) subgroups (N = 31). FT‐derived deformation indices, including left ventricular (LV) global longitudinal strain (GLS), were measured by commercial software.Statistical Tests: Mann–Whitney U‐test, Kruskal–Wallis test followed by Dunn–Bonferroni post hoc method, and receiver operating characteristic curve analysis were conducted. A P‐value of <0.05 was considered statistically significant.Results: GLS was significantly worse in TA than in controls (median [interquartile range, IQR]: TA −10.0 [−7.5 to 12.4] vs. controls −12.7 [−11.8 to 14.7]). Moreover, TA patients with LGE (+) had significantly poorer GLS than those with LGE (−) (median [IQR]: LGE (+) −6.8 [−4.0 to 8.1] vs. LGE (−) −10.7 [−8.5 to 12.9]). The reduced LVEF subgroup had significantly greater cardiac dysfunction as measured by MR‐FT than the preserved LVEF subgroup (GLS median [IQR]: reduced LVEF −7.9 [−6.2 to 11.4] vs. preserved LVEF −10.8 [−8.6 to 13.5]).Data Conclusion: Myocardial deformation impairment was found in the majority of TA patients. MR‐FT imaging may be helpful in the early diagnosis and management of TA patients.Level of Evidence: 4Technical Efficacy Stage: 5
- 28 September 2021
- Published by Wiley in Journal of Magnetic Resonance Imaging
- Vol. 55 (6) , 1828-1840
Quantification of Myocardial Deformation in Patients with Takayasu Arteritis by Cardiovascular Magnetic Resonance Feature Tracking Imaging
Background: Cardiac involvement is one of the main causes of morbidity and mortality in patients with Takayasu arteritis (TA). Early detection and intervention of cardiac damage may be helpful to reduce the mortality of TA. Magnetic resonance (MR) imaging (MRI)‐derived feature tracking (FT) is an effective quantitative method to assess myocardial deformation which may reflect early changes of cardiac function.Purpose: To explore the utility of MR‐FT as a method to detect cardiac damage in TA patients.Study Type: Retrospective.Population: Fifty‐seven TA patients who had undergone clinically indicated MRI and 57 healthy controls.Field Strength/Sequences: Balanced steady‐state free precession rest cine and 2D phase‐sensitive inversion recovery breath‐hold segmented gradient echo late gadolinium enhancement (LGE) sequences at 3.0 T.Assessment: Based on LGE images, TA patients were divided into two subgroups, LGE (+) subgroup (N = 12) and LGE (−) subgroup (N = 45). In addition, patients were further subdivided into impaired (N = 26) and preserved left ventricle ejection fraction (LVEF) subgroups (N = 31). FT‐derived deformation indices, including left ventricular (LV) global longitudinal strain (GLS), were measured by commercial software.Statistical Tests: Mann–Whitney U‐test, Kruskal–Wallis test followed by Dunn–Bonferroni post hoc method, and receiver operating characteristic curve analysis were conducted. A P‐value of <0.05 was considered statistically significant.Results: GLS was significantly worse in TA than in controls (median [interquartile range, IQR]: TA −10.0 [−7.5 to 12.4] vs. controls −12.7 [−11.8 to 14.7]). Moreover, TA patients with LGE (+) had significantly poorer GLS than those with LGE (−) (median [IQR]: LGE (+) −6.8 [−4.0 to 8.1] vs. LGE (−) −10.7 [−8.5 to 12.9]). The reduced LVEF subgroup had significantly greater cardiac dysfunction as measured by MR‐FT than the preserved LVEF subgroup (GLS median [IQR]: reduced LVEF −7.9 [−6.2 to 11.4] vs. preserved LVEF −10.8 [−8.6 to 13.5]).Data Conclusion: Myocardial deformation impairment was found in the majority of TA patients. MR‐FT imaging may be helpful in the early diagnosis and management of TA patients.Level of Evidence: 4Technical Efficacy Stage: 5