Jiansheng Li
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Jiansheng Li has a total of 30 co-authors. Here are the top co-authors by number of shared publications:
Xi Zhong
Department of Medical Imaging, Affiliated Cancer Hospital & Institute of Guangzhou Medical University, Guangzhou, China
5 shared publicationsBingui Lu
Department of Medical Imaging, Affiliated Cancer Hospital & Institute of Guangzhou Medical University, Guangzhou, People’s Republic of China
4 shared publicationsLu Huang
Department of Medical Imaging, Affiliated Cancer Hospital & Institute of Guangzhou Medical University, Guangzhou, People’s Republic of China
3 shared publicationsHui Mai
Department of Radiology, The Third Affiliated Hospital of Guangzhou Medical University, Guangzhou, People’s Republic of China
2 shared publicationsJinxue Yin
Department of Medical Imaging, Guangzhou Institute of Cancer Research, the Affiliated Cancer Hospital, Guangzhou Medical University, Guangzhou, 510095, China
2 shared publicationsLi Li
Department of Otolaryngology, The Third Affiliated Hospital of Guangzhou Medical University, Guangzhou, China
2 shared publications
Latest Publications
- Open Access
- 12 March 2025
- Published by Springer Nature in Cancer Imaging
- Vol. 25 (1) , 31
An interpretable machine learning model based on computed tomography radiomics for predicting programmed death ligand 1 expression status in gastric cancer
Background: Programmed death ligand 1 (PD-L1) expression status, closely related to immunotherapy outcomes, is a reliable biomarker for screening patients who may benefit from immunotherapy. Here, we developed and validated an interpretable machine learning (ML) model based on contrast-enhanced computed tomography (CECT) radiomics for preoperatively predicting PD-L1 expression status in patients with gastric cancer (GC). Methods: We retrospectively recruited 285 GC patients who underwent CECT and PD-L1 detection from two medical centers. A PD-L1 combined positive score (CPS) of ≥ 5 was considered to indicate a high PD-L1 expression status. Patients from center 1 were divided into training (n = 143) and validation sets (n = 62), and patients from center 2 were considered a test set (n = 80). Radiomics features were extracted from venous-phase CT images. After feature reduction and selection, 11 ML algorithms were employed to develop predictive models, and their performance in predicting PD-L1 expression status was evaluated using areas under receiver operating characteristic curves (AUCs). SHapley Additive exPlanations (SHAP) were used to interpret the optimal model and visualize the decision-making process for a single individual. Results: Nine features significantly associated with PD-L1 expression status were ultimately selected to construct the predictive model. The light gradient-boosting machine (LGBM) model demonstrated the best performance for PD-L1 high expression status prediction in the training, validation, and test sets, with AUCs of 0.841(95% CI: 0.773, 0.908), 0.834 (95% CI:0.729, 0.939), and 0.822 (95% CI: 0.718, 0.926), respectively. The SHAP summary and bar plots illustrated that a feature’s value affected the feature’s impact attributed to the model. The SHAP waterfall plots were used to visualize the decision-making process for a single individual. Conclusion: Our CT radiomics–based LGBM model may aid in preoperatively predicting PD-L1 expression status in GC patients, and the SHAP method may improve the interpretability of this model.
- 12 March 2025
- Published by Springer Nature in Cancer Imaging
- Vol. 25 (1) , 31
Open AccessAn interpretable machine learning model based on computed tomography radiomics for predicting programmed death ligand 1 expression status in gastric cancer
Background: Programmed death ligand 1 (PD-L1) expression status, closely related to immunotherapy outcomes, is a reliable biomarker for screening patients who may benefit from immunotherapy. Here, we developed and validated an interpretable machine learning (ML) model based on contrast-enhanced computed tomography (CECT) radiomics for preoperatively predicting PD-L1 expression status in patients with gastric cancer (GC). Methods: We retrospectively recruited 285 GC patients who underwent CECT and PD-L1 detection from two medical centers. A PD-L1 combined positive score (CPS) of ≥ 5 was considered to indicate a high PD-L1 expression status. Patients from center 1 were divided into training (n = 143) and validation sets (n = 62), and patients from center 2 were considered a test set (n = 80). Radiomics features were extracted from venous-phase CT images. After feature reduction and selection, 11 ML algorithms were employed to develop predictive models, and their performance in predicting PD-L1 expression status was evaluated using areas under receiver operating characteristic curves (AUCs). SHapley Additive exPlanations (SHAP) were used to interpret the optimal model and visualize the decision-making process for a single individual. Results: Nine features significantly associated with PD-L1 expression status were ultimately selected to construct the predictive model. The light gradient-boosting machine (LGBM) model demonstrated the best performance for PD-L1 high expression status prediction in the training, validation, and test sets, with AUCs of 0.841(95% CI: 0.773, 0.908), 0.834 (95% CI:0.729, 0.939), and 0.822 (95% CI: 0.718, 0.926), respectively. The SHAP summary and bar plots illustrated that a feature’s value affected the feature’s impact attributed to the model. The SHAP waterfall plots were used to visualize the decision-making process for a single individual. Conclusion: Our CT radiomics–based LGBM model may aid in preoperatively predicting PD-L1 expression status in GC patients, and the SHAP method may improve the interpretability of this model.
- Open Access
- 12 October 2023
- Published by Springer Nature in Cancer Imaging
- Vol. 23 (1) , 1-9
Reproducibility and usefulness of quantitative apparent diffusion coefficient measurements for predicting program death-ligand 1 expression in nasopharyngeal carcinoma
Background: Accurate assessment of programmed death-ligand 1 (PD-L1) expression status in nasopharyngeal carcinoma (NPC) before immunotherapy is crucial. We aimed to explore the reproducibility and usefulness of the quantitative apparent diffusion coefficient (ADC) measurements for predicting PD-L1expression status in NPC. Methods: We retrospectively recruited 134 NPC patients who underwent MRI scans and PD-L1 detection. A PD-L1 combined positive score (CPS) ≥ 20 was identified as high expression status. Patients were divide into two cohorts based on the MRI scanning devices, including a 1.5-T MRI cohort (n = 85, 44 PD-L1 high expression) and a 3.0-T MRI cohort (n = 49, 24 PD-L1 high expression). The mean ADC (ADCmean), minimum ADC (ADCmin) and maximal ADC (ADCmax) values were independently measured by two observers. The ADC measurement reproducibility was assessed by interclass correlation coefficients (ICC). The correlations between ADC parameters and CPS were analyzed by spearman’s correlation coefficient (r), and the performance for PD-L1expression status prediction was assessed by the area under receiver operating characteristic curve (AUC). Results: The measurement reproducibility of ADCmean, ADCmin and ADCmax was good in the 1.5-T MRI cohort (ICC: 0.843–0.930) and 3.0-T MRI cohort (ICC: 0.929–0.960). The ADCmean, ADCmin, and ADCmax tended to inversely correlate with the CPS (r:-0.37 - -0.52 in the 1.5-T MRI cohort, and − 0.52 - -0.60 in the 3.0-T MRI cohort; P all < 0.01). The ADCmean, ADCmin and ADCmax yielded the AUC of 0.756 (95% CI: 0.651, 0.861), 0.689 (95% CI: 0.576, 0.802), and 0.733 (95%CI: 0.626, 0.839) in the 1.5-T MRI cohort and 0.820 (95%CI: 0.703, 0.937), 0.755 (95% CI: 0.616, 0.894), and 0.760 (95%CI: 0.627, 0.893) in the 3.0-T MRI cohort for predicting PD-L1 high expression status, respectively. Conclusion: ADC measurements may act as a reproducible and feasible method to predict PD-L1 expression status in NPC.
- 12 October 2023
- Published by Springer Nature in Cancer Imaging
- Vol. 23 (1) , 1-9
Open AccessReproducibility and usefulness of quantitative apparent diffusion coefficient measurements for predicting program death-ligand 1 expression in nasopharyngeal carcinoma
Background: Accurate assessment of programmed death-ligand 1 (PD-L1) expression status in nasopharyngeal carcinoma (NPC) before immunotherapy is crucial. We aimed to explore the reproducibility and usefulness of the quantitative apparent diffusion coefficient (ADC) measurements for predicting PD-L1expression status in NPC. Methods: We retrospectively recruited 134 NPC patients who underwent MRI scans and PD-L1 detection. A PD-L1 combined positive score (CPS) ≥ 20 was identified as high expression status. Patients were divide into two cohorts based on the MRI scanning devices, including a 1.5-T MRI cohort (n = 85, 44 PD-L1 high expression) and a 3.0-T MRI cohort (n = 49, 24 PD-L1 high expression). The mean ADC (ADCmean), minimum ADC (ADCmin) and maximal ADC (ADCmax) values were independently measured by two observers. The ADC measurement reproducibility was assessed by interclass correlation coefficients (ICC). The correlations between ADC parameters and CPS were analyzed by spearman’s correlation coefficient (r), and the performance for PD-L1expression status prediction was assessed by the area under receiver operating characteristic curve (AUC). Results: The measurement reproducibility of ADCmean, ADCmin and ADCmax was good in the 1.5-T MRI cohort (ICC: 0.843–0.930) and 3.0-T MRI cohort (ICC: 0.929–0.960). The ADCmean, ADCmin, and ADCmax tended to inversely correlate with the CPS (r:-0.37 - -0.52 in the 1.5-T MRI cohort, and − 0.52 - -0.60 in the 3.0-T MRI cohort; P all < 0.01). The ADCmean, ADCmin and ADCmax yielded the AUC of 0.756 (95% CI: 0.651, 0.861), 0.689 (95% CI: 0.576, 0.802), and 0.733 (95%CI: 0.626, 0.839) in the 1.5-T MRI cohort and 0.820 (95%CI: 0.703, 0.937), 0.755 (95% CI: 0.616, 0.894), and 0.760 (95%CI: 0.627, 0.893) in the 3.0-T MRI cohort for predicting PD-L1 high expression status, respectively. Conclusion: ADC measurements may act as a reproducible and feasible method to predict PD-L1 expression status in NPC.
- Open Access
- 13 May 2022
- Published by Springer Nature in BMC Women's Health
- Vol. 22 (1) , 1-9
Clinical and MRI features of sacral insufficiency fractures after radiotherapy in patients with cervical cancer
Background: To determine the incidence, clinical and MRI features of sacral insufficiency fracture (SIF) after radiotherapy (RT) in patients with cervical cancer. Methods: Our study included 167 patients with cervical cancer after radiotherapy that underwent pelvic MRI for follow-up. MRIs included pre-enhanced T1-weighted, coronal fat-Suppressed T2-weighted (FS-T2W) and enhanced T1-weighted imaging. The clinical and MRI dates were reviewed. The gold standard of SIF was based on radiologic findings, clinical data and follow-up at least 12 months. Results: 28 patients (10.8%) with 47 sites were diagnosed with SIFs, including 9 patients with unilateral SIF and 19 patients with bilateral SIFs. The median age was 60 years (range 41–72 years), and 89.3% (25/28) of patients were postmenopausal. 64.3% (18/28) of patients were symptomatic, and 53.6% of patients (15/28) had concomitant pelvic fractures. The median interval time from RT to SIFs was 10 months (range 3–34 months). For the lesion-wise analysis based on all MR images, all lesions were detected by visualizing bone marrow edema patterns, and fracture lines were detected in 64.6% (31/47) of SIFs. No soft-tissue tumors were founded. For each MRI sequence analysis, coronal FS-T2WI detected the most bone marrow edema pattern and fracture line than T1WI or enhanced T1WI. Conclusion: SIF is a common complication in cervical cancer after radiotherapy, which has some certain clinical and MRI features. Coronal FS-T2WI may be more useful to detect and characterize these fractures than other imaging sequences.
- 13 May 2022
- Published by Springer Nature in BMC Women's Health
- Vol. 22 (1) , 1-9
Open AccessClinical and MRI features of sacral insufficiency fractures after radiotherapy in patients with cervical cancer
Background: To determine the incidence, clinical and MRI features of sacral insufficiency fracture (SIF) after radiotherapy (RT) in patients with cervical cancer. Methods: Our study included 167 patients with cervical cancer after radiotherapy that underwent pelvic MRI for follow-up. MRIs included pre-enhanced T1-weighted, coronal fat-Suppressed T2-weighted (FS-T2W) and enhanced T1-weighted imaging. The clinical and MRI dates were reviewed. The gold standard of SIF was based on radiologic findings, clinical data and follow-up at least 12 months. Results: 28 patients (10.8%) with 47 sites were diagnosed with SIFs, including 9 patients with unilateral SIF and 19 patients with bilateral SIFs. The median age was 60 years (range 41–72 years), and 89.3% (25/28) of patients were postmenopausal. 64.3% (18/28) of patients were symptomatic, and 53.6% of patients (15/28) had concomitant pelvic fractures. The median interval time from RT to SIFs was 10 months (range 3–34 months). For the lesion-wise analysis based on all MR images, all lesions were detected by visualizing bone marrow edema patterns, and fracture lines were detected in 64.6% (31/47) of SIFs. No soft-tissue tumors were founded. For each MRI sequence analysis, coronal FS-T2WI detected the most bone marrow edema pattern and fracture line than T1WI or enhanced T1WI. Conclusion: SIF is a common complication in cervical cancer after radiotherapy, which has some certain clinical and MRI features. Coronal FS-T2WI may be more useful to detect and characterize these fractures than other imaging sequences.
- Open Access
- 28 May 2021
- Published by Xia & He Publishing in Journal of Clinical and Translational Hepatology
- Vol. 10 (1) , 34-41
Added Value of Quantitative Apparent Diffusion Coefficients for Identifying Small Hepatocellular Carcinoma from Benign Nodule Categorized as LI-RADS 3 and 4 in Cirrhosis
Correct identification of small hepatocellular carcinomas (HCCs) and benign nodules in cirrhosis remains challenging, quantitative apparent diffusion coefficients (ADCs) have shown potential value in characterization of benign and malignant liver lesions. We aimed to explore the added value of ADCs in the identification of small (≤3 cm) HCCs and benign nodules categorized as Liver Imaging Reporting and Data System (LI-RADS) 3 (LR-3) and 4 (LR-4) in cirrhosis.
- 28 May 2021
- Published by Xia & He Publishing in Journal of Clinical and Translational Hepatology
- Vol. 10 (1) , 34-41
Open AccessAdded Value of Quantitative Apparent Diffusion Coefficients for Identifying Small Hepatocellular Carcinoma from Benign Nodule Categorized as LI-RADS 3 and 4 in Cirrhosis
Correct identification of small hepatocellular carcinomas (HCCs) and benign nodules in cirrhosis remains challenging, quantitative apparent diffusion coefficients (ADCs) have shown potential value in characterization of benign and malignant liver lesions. We aimed to explore the added value of ADCs in the identification of small (≤3 cm) HCCs and benign nodules categorized as Liver Imaging Reporting and Data System (LI-RADS) 3 (LR-3) and 4 (LR-4) in cirrhosis.
- Open Access
- 7 October 2020
- Published by Springer Nature in Research Square
Clinical and MRI features of sacral insufficiency fractures after radiotherapy in patients with cervical cancer
Background To determine the incidence, clinical and MRI features of sacral insufficiency fracture (SIF) after radiotherapy (RT) in patients with cervical cancer. Methods Our study included 167 patients with cervical cancer after radiotherapy that underwent pelvic MRI for follow-up. MRIs included pre-enhanced T1-weighted, coronal fat-Suppressed T2-weighted (FS-T2W) and enhanced T1-weighted imaging. The clinical and MRI dates were reviewed. The gold standard of SIF was based on radiologic findings, clinical data and follow-up at least 12 months. Results 28 patients (10.8%) with 47 sites were diagnosed with SIFs, including 9 patients with unilateral SIF and 19 patients with bilateral SIFs. The median age was 60 years (range 41–72 years), and 89.3% (25/28) of patients were postmenopausal. 64.3% (18/28) of patients were symptomatic, and 53.6% of patients (15/28) had concomitant pelvic fractures. The median interval time from RT to SIFs was 10 months (range 3 − 34 months). For the lesion-wise analysis based on all MR images, all lesions were detected by visualizing bone marrow edema patterns, and fracture lines were detected in 64.6% (31/47) of SIFs. No soft-tissue tumors were founded. For each MRI sequence analysis, coronal FS-T2WI detected the most bone marrow edema pattern and fracture line than T1WI or enhanced T1WI. Conclusion SIF is a common complication in cervical cancer after radiotherapy, which has some certain clinical and MRI features. Coronal FS-T2WI may be more useful to detect and characterize these fractures than other imaging sequences.
- 7 October 2020
- Published by Springer Nature in Research Square
Open AccessClinical and MRI features of sacral insufficiency fractures after radiotherapy in patients with cervical cancer
Background To determine the incidence, clinical and MRI features of sacral insufficiency fracture (SIF) after radiotherapy (RT) in patients with cervical cancer. Methods Our study included 167 patients with cervical cancer after radiotherapy that underwent pelvic MRI for follow-up. MRIs included pre-enhanced T1-weighted, coronal fat-Suppressed T2-weighted (FS-T2W) and enhanced T1-weighted imaging. The clinical and MRI dates were reviewed. The gold standard of SIF was based on radiologic findings, clinical data and follow-up at least 12 months. Results 28 patients (10.8%) with 47 sites were diagnosed with SIFs, including 9 patients with unilateral SIF and 19 patients with bilateral SIFs. The median age was 60 years (range 41–72 years), and 89.3% (25/28) of patients were postmenopausal. 64.3% (18/28) of patients were symptomatic, and 53.6% of patients (15/28) had concomitant pelvic fractures. The median interval time from RT to SIFs was 10 months (range 3 − 34 months). For the lesion-wise analysis based on all MR images, all lesions were detected by visualizing bone marrow edema patterns, and fracture lines were detected in 64.6% (31/47) of SIFs. No soft-tissue tumors were founded. For each MRI sequence analysis, coronal FS-T2WI detected the most bone marrow edema pattern and fracture line than T1WI or enhanced T1WI. Conclusion SIF is a common complication in cervical cancer after radiotherapy, which has some certain clinical and MRI features. Coronal FS-T2WI may be more useful to detect and characterize these fractures than other imaging sequences.
- Open Access
- 24 January 2020
- Published by Frontiers Media SA in Frontiers in Oncology
- Vol. 10, 15
Differentiation of Cervical Spine Osteoradionecrosis and Bone Metastasis After Radiotherapy Detected by Bone Scan in Patients With Nasopharyngeal Carcinoma: Role of Magnetic Resonance Imaging
Background: Osteoradionecrosis (ORN) of the cervical spine is a serious complication after radiotherapy (RT), which may show increased radiotracer uptake on a bone scan (BS) and be mistaken as metastasis. We aimed to assess the value of magnetic resonance imaging (MRI) in the differentiation of cervical spine ORN from bone metastasis after RT detected by BS in nasopharyngeal carcinoma (NPC). Methods: In this retrospective study, 35 NPC patients who had undergone RT were enrolled, of whom 21 patients showed cervical spine ORN and 14 showed bone metastasis. New areas of increased radiotracer uptake in the cervical spine on a BS were noted in all patients, following which the patients underwent neck MRI for further assessment. Two radiologists independently reviewed two sets of images including a BS set and an MRI set (MRI with BS) and reached a consensus. The diagnostic sensitivity, specificity, and accuracy for ORN detection were calculated, and interobserver agreement was evaluated using the kappa test. Results: A total of 75 cervical spine lesions were identified (44, ORN; 31 metastases). The BS set analysis showed that the diagnostic sensitivity, specificity, and accuracy were only 38.6, 48.3, and 42.7%, respectively, for differentiation of cervical spine ORN from bone metastasis. On the other hand, the MRI set analysis showed that the diagnostic sensitivity, specificity, and accuracy increased to 86.4, 90.3, and 88.0%, respectively. The interobserver agreement for the MRI set was determined to be very good (κ = 0.92). Conclusion: MRI is a reliable technique for the further discrimination of emerging cervical spine lesions after RT detected by BS. Furthermore, it could be a better differential diagnosis technique for distinguishing ORN from metastasis and may help avoid a wrong assignment of the patient to a metastatic stage with indication for treatment with supplemental toxicity and a subsequent palliative strategy.
- 24 January 2020
- Published by Frontiers Media SA in Frontiers in Oncology
- Vol. 10, 15
Open AccessDifferentiation of Cervical Spine Osteoradionecrosis and Bone Metastasis After Radiotherapy Detected by Bone Scan in Patients With Nasopharyngeal Carcinoma: Role of Magnetic Resonance Imaging
Background: Osteoradionecrosis (ORN) of the cervical spine is a serious complication after radiotherapy (RT), which may show increased radiotracer uptake on a bone scan (BS) and be mistaken as metastasis. We aimed to assess the value of magnetic resonance imaging (MRI) in the differentiation of cervical spine ORN from bone metastasis after RT detected by BS in nasopharyngeal carcinoma (NPC). Methods: In this retrospective study, 35 NPC patients who had undergone RT were enrolled, of whom 21 patients showed cervical spine ORN and 14 showed bone metastasis. New areas of increased radiotracer uptake in the cervical spine on a BS were noted in all patients, following which the patients underwent neck MRI for further assessment. Two radiologists independently reviewed two sets of images including a BS set and an MRI set (MRI with BS) and reached a consensus. The diagnostic sensitivity, specificity, and accuracy for ORN detection were calculated, and interobserver agreement was evaluated using the kappa test. Results: A total of 75 cervical spine lesions were identified (44, ORN; 31 metastases). The BS set analysis showed that the diagnostic sensitivity, specificity, and accuracy were only 38.6, 48.3, and 42.7%, respectively, for differentiation of cervical spine ORN from bone metastasis. On the other hand, the MRI set analysis showed that the diagnostic sensitivity, specificity, and accuracy increased to 86.4, 90.3, and 88.0%, respectively. The interobserver agreement for the MRI set was determined to be very good (κ = 0.92). Conclusion: MRI is a reliable technique for the further discrimination of emerging cervical spine lesions after RT detected by BS. Furthermore, it could be a better differential diagnosis technique for distinguishing ORN from metastasis and may help avoid a wrong assignment of the patient to a metastatic stage with indication for treatment with supplemental toxicity and a subsequent palliative strategy.