Kai Li
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
Kai Li has a total of 300 co-authors. Here are the top co-authors by number of shared publications:
Qingjing Zeng
Department of Ultrasound, The Third Affliated Hospital of Sun Yat-sen University, Guangzhou City, Guangdong Province, China
15 shared publicationsRongqin Zheng
Department of Ultrasound, The Third Affiliated Hospital of Sun Yat-sen University, Guangzhou, China
10 shared publicationsLei Tan
Department of Ultrasound, Guangdong Key Laboratory of Liver Disease Research, The Third Affiliated Hospital of Sun Yat-sen University, 600 Tianhe Road, Guangzhou 510630, China
8 shared publicationsQiannan Huang
Department of Ultrasound, The Third Affiliated Hospital of Sun Yat-sen University, Guangzhou, China
8 shared publications- Ping Liang
Department of Interventional Ultrasound, Chinese PLA General Hospital, 28 Fuxing Road, Beijing 100853, China
7 shared publications - Rongqin Zheng
The Third Affiliated Hospital of Sun Yat-sen University, Guangzhou, China
7 shared publications
Latest Publications
- Open Access
- 27 October 2026
- Published by S. Karger AG in Liver Cancer
2026 International Expert Consensus Recommendations on Image-Guided Ablation for Colorectal Liver Metastases
Image-guided ablation has emerged as a potentially curative treatment for colorectal cancer liver metastases (CRLM), yet substantial heterogeneity persists in patient selection, technical execution, and perioperative management. To establish standardized, evidence-based guidance, a multidisciplinary panel of 68 experts including image-guided interventionalists, surgeons, and oncologists conducted a systematic review of literature from 2000 to 2025 and developed consensus recommendations through a modified two-round Delphi process. Evidence was graded using the Oxford Centre for Evidence-Based Medicine framework, achieving at least conditional consensus (≥75% agreement) for all statements. Finally, twenty consensus recommendations were established addressing patient selection, imaging evaluation, image-guided techniques, ablation strategies, complication management, surveillance protocols, and combination therapies. This expert consensus provides a comprehensive framework for optimizing image-guided ablation as an integral component of multidisciplinary, personalized CRLM management.
- 27 October 2026
- Published by S. Karger AG in Liver Cancer
Open Access2026 International Expert Consensus Recommendations on Image-Guided Ablation for Colorectal Liver Metastases
Image-guided ablation has emerged as a potentially curative treatment for colorectal cancer liver metastases (CRLM), yet substantial heterogeneity persists in patient selection, technical execution, and perioperative management. To establish standardized, evidence-based guidance, a multidisciplinary panel of 68 experts including image-guided interventionalists, surgeons, and oncologists conducted a systematic review of literature from 2000 to 2025 and developed consensus recommendations through a modified two-round Delphi process. Evidence was graded using the Oxford Centre for Evidence-Based Medicine framework, achieving at least conditional consensus (≥75% agreement) for all statements. Finally, twenty consensus recommendations were established addressing patient selection, imaging evaluation, image-guided techniques, ablation strategies, complication management, surveillance protocols, and combination therapies. This expert consensus provides a comprehensive framework for optimizing image-guided ablation as an integral component of multidisciplinary, personalized CRLM management.
- 31 August 2026
- Published by Springer Nature in Journal of Ultrasound
- Vol. 29 (3) , 707-715
The ultrasound and contrast-enhanced ultrasound characteristics of IgG4-related cholangitis: compared with cholangiocarcinoma
- 31 August 2026
- Published by Springer Nature in Journal of Ultrasound
- Vol. 29 (3) , 707-715
The ultrasound and contrast-enhanced ultrasound characteristics of IgG4-related cholangitis: compared with cholangiocarcinoma
- 1 June 2026
- Published by Radiological Society of North America (RSNA) in Radiology
- Vol. 319 (3) , e251003
Sequencing Ablation and Systemic Therapy in Colorectal Liver Oligometastases: An Upfront versus Delayed Approach Nationwide Analysis
In this multicenter cohort, upfront ablation plus systemic therapy significantly improved both progression-free survival and overall survival in patients with colorectal liver oligometastases compared with delayed ablation.
- 1 June 2026
- Published by Radiological Society of North America (RSNA) in Radiology
- Vol. 319 (3) , e251003
Sequencing Ablation and Systemic Therapy in Colorectal Liver Oligometastases: An Upfront versus Delayed Approach Nationwide Analysis
In this multicenter cohort, upfront ablation plus systemic therapy significantly improved both progression-free survival and overall survival in patients with colorectal liver oligometastases compared with delayed ablation.
- 29 November 2025
- Published by Elsevier in Journal of Controlled Release
- Vol. 390, 114480
Nanovesicles integrating PD-1-mediated targeting and CRISPR/Cas9-based CD47 editing for dual immune checkpoint blockade
- 29 November 2025
- Published by Elsevier in Journal of Controlled Release
- Vol. 390, 114480
Nanovesicles integrating PD-1-mediated targeting and CRISPR/Cas9-based CD47 editing for dual immune checkpoint blockade
- Open Access
- 14 November 2025
- Published by Springer Nature in Research Square
Evaluation of Learning Curves for US-CT/MR and US-US Fusion Imaging in the Liver
The aim of this retrospective study was to investigate the learning curve of ultrasound (US)–computed tomography (CT)/magnetic resonance (MR) as well as US–three-dimensional (3D) US fusion imaging for liver tumors. Patients intending to undergo liver tumor thermal ablation in our institution were enrolled. Two operators were investigated, one senior and one junior US doctor. Both operators practiced on the same patients for each US fusion imaging technique, respectively. The results of registration were blindly judged by two senior sonographers with > 5-year experience on US fusion imaging. The registration time and success rate of registration were recorded. The trend of registration time with practice attempts was analyzed using CUSUM analysis. Both operators reached the criteria to terminate the learning in both two types of fusion imaging after 15 practices. The CUSUM analysis of the registration time showed that phases A and B were divided by the 10–11th practices for US–CT/MR fusion imaging and 5th practices for US–3D US fusion imaging for both operators. In the US–CT/MR fusion imaging, the success rates of registration for the senior operator in both phases were high (91–100%), whereas that for the junior operator in the phase A was low (50%). By contrast, in US–US fusion imaging, the success rates of registration for the senior and the junior operator in both phases were high (100%). In conclusion, US–3D US fusion imaging has a shorter learning curve than US–CT/MR fusion imaging and is less dependent on experience.
- 14 November 2025
- Published by Springer Nature in Research Square
Open AccessEvaluation of Learning Curves for US-CT/MR and US-US Fusion Imaging in the Liver
The aim of this retrospective study was to investigate the learning curve of ultrasound (US)–computed tomography (CT)/magnetic resonance (MR) as well as US–three-dimensional (3D) US fusion imaging for liver tumors. Patients intending to undergo liver tumor thermal ablation in our institution were enrolled. Two operators were investigated, one senior and one junior US doctor. Both operators practiced on the same patients for each US fusion imaging technique, respectively. The results of registration were blindly judged by two senior sonographers with > 5-year experience on US fusion imaging. The registration time and success rate of registration were recorded. The trend of registration time with practice attempts was analyzed using CUSUM analysis. Both operators reached the criteria to terminate the learning in both two types of fusion imaging after 15 practices. The CUSUM analysis of the registration time showed that phases A and B were divided by the 10–11th practices for US–CT/MR fusion imaging and 5th practices for US–3D US fusion imaging for both operators. In the US–CT/MR fusion imaging, the success rates of registration for the senior operator in both phases were high (91–100%), whereas that for the junior operator in the phase A was low (50%). By contrast, in US–US fusion imaging, the success rates of registration for the senior and the junior operator in both phases were high (100%). In conclusion, US–3D US fusion imaging has a shorter learning curve than US–CT/MR fusion imaging and is less dependent on experience.
- Open Access
- 1 May 2025
- Published by Springer Nature in Nature Communications
- Vol. 16 (1) , 1-14
Vision transformer-based model can optimize curative-intent treatment for patients with recurrent hepatocellular carcinoma
The treatment selection for recurrent hepatocellular carcinoma (rHCC) within Milan criteria after hepatectomy remains challenging. Here, we present HEROVision, a Vision Transformer-based model designed for personalized prognosis prediction and treatment optimization between thermal ablation (TA) and surgical resection (SR). HEROVision is trained on initial HCC cohorts (8492 images; 772 patients) and independently tested on rHCC cohorts (9163 images; 833 patients) from five centers. Propensity score matching (PSM) forms two groups of rHCC patients underwent TA and SR to fairly evaluate whether optimized treatment selection by HEROVision have clinical benefits. HEROVision significantly outperforms all six guideline staging systems in the external testing cohort, both in time-dependent concordance index and area under the curve (all P < 0.002). After PSM, 35.9% (23/64) and 6.6% (6/91) high-risk rHCC patients are identified, who could achieve improved prognosis by changing their treatments. HEROVision shows promise in optimizing individualized treatment between TA and SR for early-stage rHCC, complementing current clinical guidelines.
- 1 May 2025
- Published by Springer Nature in Nature Communications
- Vol. 16 (1) , 1-14
Open AccessVision transformer-based model can optimize curative-intent treatment for patients with recurrent hepatocellular carcinoma
The treatment selection for recurrent hepatocellular carcinoma (rHCC) within Milan criteria after hepatectomy remains challenging. Here, we present HEROVision, a Vision Transformer-based model designed for personalized prognosis prediction and treatment optimization between thermal ablation (TA) and surgical resection (SR). HEROVision is trained on initial HCC cohorts (8492 images; 772 patients) and independently tested on rHCC cohorts (9163 images; 833 patients) from five centers. Propensity score matching (PSM) forms two groups of rHCC patients underwent TA and SR to fairly evaluate whether optimized treatment selection by HEROVision have clinical benefits. HEROVision significantly outperforms all six guideline staging systems in the external testing cohort, both in time-dependent concordance index and area under the curve (all P < 0.002). After PSM, 35.9% (23/64) and 6.6% (6/91) high-risk rHCC patients are identified, who could achieve improved prognosis by changing their treatments. HEROVision shows promise in optimizing individualized treatment between TA and SR for early-stage rHCC, complementing current clinical guidelines.
- Open Access
- 10 March 2025
- Published by Oxford University Press (OUP) in The Oncologist
- Vol. 30 (3)
Predicting early recurrence of hepatocellular carcinoma after thermal ablation based on longitudinal MRI with a deep learning approach
Background: Accurate prediction of early recurrence (ER) is essential to improve the prognosis of patients with hepatocellular carcinoma (HCC) underwent thermal ablation (TA). Therefore, a deep learning model system using longitudinal magnetic resonance imaging (MRI) was developed to predict ER of patients with HCC. Methods: From 2014, April to 2017, May, a total of 289 eligible patients with HCC underwent TA were retrospectively enrolled from 3 hospitals and assigned into one training cohort (n = 254) and one external testing cohort (n = 35). Two deep learning models (Pre and PrePost) were developed using the pre-operative MRI and longitudinal MRI (pre- and post-operative) to predict ER for the patients with HCC after TA, respectively. Then, an integrated model (DL_Clinical) incorporating PrePost model signature and clinical variables was built for post-ablation ER risk stratification for the patients with HCC. Results: In the external testing cohort, the area under the receiver operating characteristic curve (AUC) of the DL_Clinical model was better than that of the Clinical (0.740 vs 0.571), Pre (0.740 vs 0.648), and PrePost model (0.740 vs 0.689). Additionally, there was a significant difference in RFS between the high- and low-risk groups which were divided by the DL_Clinical model (P = .04). Conclusions: The PrePost model developed using longitudinal MRI showed outstanding performance for predicting post-ablation ER of HCC. The DL_Clinical model could stratify the patients into high- and low-risk groups, which may help physicians in treatment and surveillance strategy selection in clinical practice.
- 10 March 2025
- Published by Oxford University Press (OUP) in The Oncologist
- Vol. 30 (3)
Open AccessPredicting early recurrence of hepatocellular carcinoma after thermal ablation based on longitudinal MRI with a deep learning approach
Background: Accurate prediction of early recurrence (ER) is essential to improve the prognosis of patients with hepatocellular carcinoma (HCC) underwent thermal ablation (TA). Therefore, a deep learning model system using longitudinal magnetic resonance imaging (MRI) was developed to predict ER of patients with HCC. Methods: From 2014, April to 2017, May, a total of 289 eligible patients with HCC underwent TA were retrospectively enrolled from 3 hospitals and assigned into one training cohort (n = 254) and one external testing cohort (n = 35). Two deep learning models (Pre and PrePost) were developed using the pre-operative MRI and longitudinal MRI (pre- and post-operative) to predict ER for the patients with HCC after TA, respectively. Then, an integrated model (DL_Clinical) incorporating PrePost model signature and clinical variables was built for post-ablation ER risk stratification for the patients with HCC. Results: In the external testing cohort, the area under the receiver operating characteristic curve (AUC) of the DL_Clinical model was better than that of the Clinical (0.740 vs 0.571), Pre (0.740 vs 0.648), and PrePost model (0.740 vs 0.689). Additionally, there was a significant difference in RFS between the high- and low-risk groups which were divided by the DL_Clinical model (P = .04). Conclusions: The PrePost model developed using longitudinal MRI showed outstanding performance for predicting post-ablation ER of HCC. The DL_Clinical model could stratify the patients into high- and low-risk groups, which may help physicians in treatment and surveillance strategy selection in clinical practice.
- Open Access
- 1 January 2025
- Published by XMLink in Korean Journal of Radiology
- Vol. 26 (4) , 346-359
Combination of CT/MRI LI-RADS With Second-Line Contrast-Enhanced Ultrasound Using Sulfur Hexafluoride or Perfluorobutane for Diagnosing Hepatocellular Carcinoma in High-Risk Patients
Li Y, et al. Korean J Radiol. 2025 Feb;26:e29. https://doi.org/10.3348/kjr.2024.0980
- 1 January 2025
- Published by XMLink in Korean Journal of Radiology
- Vol. 26 (4) , 346-359
Open AccessCombination of CT/MRI LI-RADS With Second-Line Contrast-Enhanced Ultrasound Using Sulfur Hexafluoride or Perfluorobutane for Diagnosing Hepatocellular Carcinoma in High-Risk Patients
Li Y, et al. Korean J Radiol. 2025 Feb;26:e29. https://doi.org/10.3348/kjr.2024.0980
- Open Access
- 1 January 2025
- Published by SAGE Publications in Therapeutic Advances in Medical Oncology
- Vol. 17
Surgical resection versus thermal ablation: comparison of sequential options after successful TACE downstaging therapy for unresectable hepatocellular carcinoma
Background: Transarterial chemoembolization (TACE) is an effective and safe downstaging therapy for hepatocellular carcinoma (HCC). However, the selection of sequential therapeutic modalities is still controversial. Objectives: This study compared the effectiveness and safety of surgical resection (SR) and thermal ablation (TA) after patients with HCC underwent TACE downstaging therapy. Design: A retrospective, multi-institutional study. Methods: From June 2008 to October 2022, a total of 4782 consecutive patients with HCC beyond the initial Milan criteria underwent TACE at 12 hospitals. Among them, 609 patients who received successful downstaging therapy were retrospectively reviewed. Among them, 209 patients underwent an SR, and 390 patients received TA after TACE. The propensity score matching (PSM) method was applied to reduce selection bias between groups. Cumulative overall survival (OS) and progression-free survival (PFS) were compared using the Kaplan–Meier method with the log-rank test. Results: After PSM 1:1 ( n = 185 in both groups), the cumulative 1-, 3-, 5-, and 10-year OS rates were 98.8%, 89.3%, 82.9%, and 64.4%, respectively, in the SR group and 99.5%, 88.4%, 75.3%, and 53.9%, respectively, in the TA group; these two groups were not significantly different (HR: 1.22; 95% CI: 0.78–1.89; p = 0.381). The cumulative 1-, 3-, 5-, and 10-year PFS rates were 88.5%, 69.2%, 58.8%, and 32.2%, respectively, in the SR group and 90.6%, 71.4%, 53.1%, and 32.0%, respectively, in the TA group, revealing no significant difference between the two groups (HR: 0.97; 95% CI: 0.71–1.32; p = 0.855). Conclusion: For HCC patients beyond the Milan criteria who received TACE downstaging therapy, TA might be acceptable as an alternative to SR in the first-line sequential treatment scheme.
- 1 January 2025
- Published by SAGE Publications in Therapeutic Advances in Medical Oncology
- Vol. 17
Open AccessSurgical resection versus thermal ablation: comparison of sequential options after successful TACE downstaging therapy for unresectable hepatocellular carcinoma
Background: Transarterial chemoembolization (TACE) is an effective and safe downstaging therapy for hepatocellular carcinoma (HCC). However, the selection of sequential therapeutic modalities is still controversial. Objectives: This study compared the effectiveness and safety of surgical resection (SR) and thermal ablation (TA) after patients with HCC underwent TACE downstaging therapy. Design: A retrospective, multi-institutional study. Methods: From June 2008 to October 2022, a total of 4782 consecutive patients with HCC beyond the initial Milan criteria underwent TACE at 12 hospitals. Among them, 609 patients who received successful downstaging therapy were retrospectively reviewed. Among them, 209 patients underwent an SR, and 390 patients received TA after TACE. The propensity score matching (PSM) method was applied to reduce selection bias between groups. Cumulative overall survival (OS) and progression-free survival (PFS) were compared using the Kaplan–Meier method with the log-rank test. Results: After PSM 1:1 ( n = 185 in both groups), the cumulative 1-, 3-, 5-, and 10-year OS rates were 98.8%, 89.3%, 82.9%, and 64.4%, respectively, in the SR group and 99.5%, 88.4%, 75.3%, and 53.9%, respectively, in the TA group; these two groups were not significantly different (HR: 1.22; 95% CI: 0.78–1.89; p = 0.381). The cumulative 1-, 3-, 5-, and 10-year PFS rates were 88.5%, 69.2%, 58.8%, and 32.2%, respectively, in the SR group and 90.6%, 71.4%, 53.1%, and 32.0%, respectively, in the TA group, revealing no significant difference between the two groups (HR: 0.97; 95% CI: 0.71–1.32; p = 0.855). Conclusion: For HCC patients beyond the Milan criteria who received TACE downstaging therapy, TA might be acceptable as an alternative to SR in the first-line sequential treatment scheme.
- Open Access
- 1 December 2024
- Published by Wolters Kluwer Health in Journal of Cancer Research and Therapeutics
- Vol. 20 (7) , 2103-2109
Factors retarding enhanced recovery from thermal ablation of liver tumors: A single-center prospective study
Purpose: To evaluate the risk factors that may delay enhanced recovery in the ablation of liver tumors. Methods: A total of 310 patients who underwent ultrasound-guided ablation of liver tumors under general anesthesia were prospectively enrolled. Baseline data, intraoperative parameters, and postoperative events were evaluated. Postoperative pain was scored using the visual analog scale (VAS). Logistic regression analysis was conducted for univariate and multivariate analyses. Results: The study included 42 females (13.5%) and 268 males (86.5%). The mean age of the sample was 57 ± 11 years old. The average length of stay (LOS) was 4.3 ± 2.4 days. A total of 199 out of 310 patients (64.2%) experienced moderate to severe pain (VAS score > 3). Seventy out of 310 patients (22.6%) experienced other complications. In the multivariable analysis, the number of lesions [odds ratio (95% confidence interval): 3.23 (2.15–4.84); P < 0.001], maximum diameter of lesions [1.12 (1.07–1.17), P < 0.001], and smallest distance between the lesions and the liver capsule [0.91 (0.89–0.94), P < 0.001] were risk factors for postoperative pain (VAS > 3). A history of alcohol consumption [2.62 (1.33–5.19), P = 0.005] was a risk factor for other complications. Surgical history [0.40 (0.24–0.67), P = 0.001] was a protective factor against LOS. Total operation time [1.01, 1.00–1.01, P = 0.009] was a mild risk factor for LOS. Conclusion: The number of lesions, maximum diameter of the lesions, smallest distance between the lesions and the liver capsule, total operation time, and a history of alcohol use were risk factors that may delay enhanced recovery in patients undergoing ablation of liver tumors. These findings may provide evidence to promote the use of the enhanced recovery after surgery protocol.
- 1 December 2024
- Published by Wolters Kluwer Health in Journal of Cancer Research and Therapeutics
- Vol. 20 (7) , 2103-2109
Open AccessFactors retarding enhanced recovery from thermal ablation of liver tumors: A single-center prospective study
Purpose: To evaluate the risk factors that may delay enhanced recovery in the ablation of liver tumors. Methods: A total of 310 patients who underwent ultrasound-guided ablation of liver tumors under general anesthesia were prospectively enrolled. Baseline data, intraoperative parameters, and postoperative events were evaluated. Postoperative pain was scored using the visual analog scale (VAS). Logistic regression analysis was conducted for univariate and multivariate analyses. Results: The study included 42 females (13.5%) and 268 males (86.5%). The mean age of the sample was 57 ± 11 years old. The average length of stay (LOS) was 4.3 ± 2.4 days. A total of 199 out of 310 patients (64.2%) experienced moderate to severe pain (VAS score > 3). Seventy out of 310 patients (22.6%) experienced other complications. In the multivariable analysis, the number of lesions [odds ratio (95% confidence interval): 3.23 (2.15–4.84); P < 0.001], maximum diameter of lesions [1.12 (1.07–1.17), P < 0.001], and smallest distance between the lesions and the liver capsule [0.91 (0.89–0.94), P < 0.001] were risk factors for postoperative pain (VAS > 3). A history of alcohol consumption [2.62 (1.33–5.19), P = 0.005] was a risk factor for other complications. Surgical history [0.40 (0.24–0.67), P = 0.001] was a protective factor against LOS. Total operation time [1.01, 1.00–1.01, P = 0.009] was a mild risk factor for LOS. Conclusion: The number of lesions, maximum diameter of the lesions, smallest distance between the lesions and the liver capsule, total operation time, and a history of alcohol use were risk factors that may delay enhanced recovery in patients undergoing ablation of liver tumors. These findings may provide evidence to promote the use of the enhanced recovery after surgery protocol.