Ting Chen
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Ting Chen has a total of 38 co-authors. Here are the top co-authors by number of shared publications:
- Fei Hou
Department of Nuclear Medicine, Yunnan Cancer Hospital, The Third Affiliated Hospital of Kunming Medical University, Peking University Cancer Hospital Yunnan, Kunming, China
36 shared publications - Haolin Cai
Department of Nuclear Medicine, Yunnan Cancer Hospital, The Third Affiliated Hospital of Kunming Medical University, Peking University Cancer Hospital Yunnan, Kunming, China
36 shared publications - Hongbo Zhao
Department of Laboratory Animal Science, Kunming Medical University, Kunming, China
36 shared publications - Lin Lu
Institute of Biomedical Engineering Kunming Medical University Kunming Yunnan China
36 shared publications - Xiaoqi Peng
Department of Nuclear Medicine, Yunnan Cancer Hospital, The Third Affiliated Hospital of Kunming Medical University, Peking University Cancer Hospital Yunnan, Kunming, China
36 shared publications - Yinghui Wang
Department of Nuclear Medicine, Yunnan Cancer Hospital, The Third Affiliated Hospital of Kunming Medical University, Peking University Cancer Hospital Yunnan, Kunming, China
36 shared publications
Latest Publications
- Open Access
- 13 August 2026
- Published by Elsevier in SSRN Electronic Journal
Preoperative CT-Based Deep Learning for Predicting Distant Metastasis in Differentiated Thyroid Carcinoma: A Two-Center External Validation Study
Background: Preoperative prediction of distant metastasis (M1) in differentiated thyroid carcinoma (DTC) remains challenging, although it is critical for staging, risk stratification, and treatment planning. We aimed to develop and externally validate a preoperative CT-based deep-learning model for predicting M1 status in DTC.Methods: In this retrospective two-center study, patients with DTC treated between January 2013 and May 2023 who underwent preoperative contrast-enhanced neck CT were included. Center 1 was used for model development and internal cross-validation (n=881; M1, n=119), and Center 2 was used for independent external validation (n=145; M1, n=15). ThyroidFusionNet evaluated candidate configurations integrating CT images with clinicopathological variables, radiomics, and tumor-habitat features through systematic ablation analysis. Exploratory post-hoc interpretability analyses included CT attribution, radiomics attribution, and WSI-based pathology anchoring.Findings: Among prespecified configurations, the CT+radiomics model achieved the best overall performance. It achieved an internal AUC of 0.889 (95% CI 0.860–0.917) and an external AUC of 0.952 (95% CI 0.910–0.983). At the predefined threshold, external sensitivity, specificity, PPV, and NPV were 0.933, 0.846, 0.412, and 0.991, respectively. More complex multimodal fusion strategies did not improve external validation performance. Interpretability analyses suggested that model predictions were associated with suspicious lymph-node regions, distributed radiomics features, and exploratory histomorphologic prototypes.Interpretation: A preoperative CT+radiomics deep-learning model demonstrated strong discrimination for predicting distant metastasis status in DTC using routinely acquired CT data. The model may serve as a risk-stratification tool to predict patients who require more intensive metastatic evaluation. Prospective multicenter validation is warranted before clinical implementation.
- 13 August 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessPreoperative CT-Based Deep Learning for Predicting Distant Metastasis in Differentiated Thyroid Carcinoma: A Two-Center External Validation Study
Background: Preoperative prediction of distant metastasis (M1) in differentiated thyroid carcinoma (DTC) remains challenging, although it is critical for staging, risk stratification, and treatment planning. We aimed to develop and externally validate a preoperative CT-based deep-learning model for predicting M1 status in DTC.Methods: In this retrospective two-center study, patients with DTC treated between January 2013 and May 2023 who underwent preoperative contrast-enhanced neck CT were included. Center 1 was used for model development and internal cross-validation (n=881; M1, n=119), and Center 2 was used for independent external validation (n=145; M1, n=15). ThyroidFusionNet evaluated candidate configurations integrating CT images with clinicopathological variables, radiomics, and tumor-habitat features through systematic ablation analysis. Exploratory post-hoc interpretability analyses included CT attribution, radiomics attribution, and WSI-based pathology anchoring.Findings: Among prespecified configurations, the CT+radiomics model achieved the best overall performance. It achieved an internal AUC of 0.889 (95% CI 0.860–0.917) and an external AUC of 0.952 (95% CI 0.910–0.983). At the predefined threshold, external sensitivity, specificity, PPV, and NPV were 0.933, 0.846, 0.412, and 0.991, respectively. More complex multimodal fusion strategies did not improve external validation performance. Interpretability analyses suggested that model predictions were associated with suspicious lymph-node regions, distributed radiomics features, and exploratory histomorphologic prototypes.Interpretation: A preoperative CT+radiomics deep-learning model demonstrated strong discrimination for predicting distant metastasis status in DTC using routinely acquired CT data. The model may serve as a risk-stratification tool to predict patients who require more intensive metastatic evaluation. Prospective multicenter validation is warranted before clinical implementation.
- Open Access
- 16 May 2026
- Published by Springer Nature in The International Journal of Cardiovascular Imaging
- Vol. 42 (9) , 1709-1721
Evaluation of the diagnostic value of dual-time-point, multi-parameter 99mTc-PYP imaging in diagnosing Transthyretin Cardiac Amyloidosis
Objective To assess the clinical value of dual-time-point planar imaging and SPECT/CT tomographic imaging of 99mTc-PYP in diagnosing transthyretin cardiac amyloidosis (ATTR-CA), and to compare the diagnostic performance of different imaging time points and methods. Methods A retrospective analysis of 98 suspected CA patients (ATTR-CA 12; AL-CA 23; non-CA 63). Early imaging at 1 h and delayed imaging at 3 h included planar and tomographic imaging. Evaluations used the Perugini visual grading, heart-to-contralateral lung uptake ratio (H/CL), and tomographic semi-quantitative scoring. Pathology biopsy and genetic testing served as the gold standard. Sensitivity, specificity, and accuracy were compared across methods. Results Delayed tomographic imaging for ATTR-CA achieved a sensitivity of 100.0%, specificity of 96.5%, and accuracy of 96.9%, the best among all methods. The earliest Perugini scoring method had the lowest accuracy (80.6%), and early imaging using the H/CL method had the lowest diagnostic sensitivity (58.3%). SPECT/CT imaging at 3 h provides significant incremental diagnostic value compared to planar imaging and offers the greatest benefits, Cases that were missed during the initial review have been reclassified.Delayed imaging was superior to early imaging overall, and tomographic imaging significantly reduced false positives caused by blood-pool artifacts. Conclusion Dual-time-point 99mTc-PYP imaging with tomographic imaging can significantly improve the diagnostic accuracy for ATTR-CA, with delayed SPECT/CT performing best. For cases with atypical early imaging or diagnostic difficulty due to blood-pool tracer distribution, when Perugini grade is 2, early H/CL ≤ 1.5 or delayed H/CL ≤ 1.3, or in highly suspected cases where early imaging provides insufficient diagnostic evidence, routine inclusion of delayed SPECT/CT in the imaging workflow for suspected ATTR-CA is recommended to optimize subtyping and guide subsequent therapy.
- 16 May 2026
- Published by Springer Nature in The International Journal of Cardiovascular Imaging
- Vol. 42 (9) , 1709-1721
Open AccessEvaluation of the diagnostic value of dual-time-point, multi-parameter 99mTc-PYP imaging in diagnosing Transthyretin Cardiac Amyloidosis
Objective To assess the clinical value of dual-time-point planar imaging and SPECT/CT tomographic imaging of 99mTc-PYP in diagnosing transthyretin cardiac amyloidosis (ATTR-CA), and to compare the diagnostic performance of different imaging time points and methods. Methods A retrospective analysis of 98 suspected CA patients (ATTR-CA 12; AL-CA 23; non-CA 63). Early imaging at 1 h and delayed imaging at 3 h included planar and tomographic imaging. Evaluations used the Perugini visual grading, heart-to-contralateral lung uptake ratio (H/CL), and tomographic semi-quantitative scoring. Pathology biopsy and genetic testing served as the gold standard. Sensitivity, specificity, and accuracy were compared across methods. Results Delayed tomographic imaging for ATTR-CA achieved a sensitivity of 100.0%, specificity of 96.5%, and accuracy of 96.9%, the best among all methods. The earliest Perugini scoring method had the lowest accuracy (80.6%), and early imaging using the H/CL method had the lowest diagnostic sensitivity (58.3%). SPECT/CT imaging at 3 h provides significant incremental diagnostic value compared to planar imaging and offers the greatest benefits, Cases that were missed during the initial review have been reclassified.Delayed imaging was superior to early imaging overall, and tomographic imaging significantly reduced false positives caused by blood-pool artifacts. Conclusion Dual-time-point 99mTc-PYP imaging with tomographic imaging can significantly improve the diagnostic accuracy for ATTR-CA, with delayed SPECT/CT performing best. For cases with atypical early imaging or diagnostic difficulty due to blood-pool tracer distribution, when Perugini grade is 2, early H/CL ≤ 1.5 or delayed H/CL ≤ 1.3, or in highly suspected cases where early imaging provides insufficient diagnostic evidence, routine inclusion of delayed SPECT/CT in the imaging workflow for suspected ATTR-CA is recommended to optimize subtyping and guide subsequent therapy.
- Top Cited PapersOpen Access
- 31 October 2024
- Published by Elsevier in eClinicalMedicine
- Vol. 77, 102913
Development and validation of an interpretable machine learning model for predicting the risk of distant metastasis in papillary thyroid cancer: a multicenter study
- 31 October 2024
- Published by Elsevier in eClinicalMedicine
- Vol. 77, 102913
Top Cited PapersOpen AccessDevelopment and validation of an interpretable machine learning model for predicting the risk of distant metastasis in papillary thyroid cancer: a multicenter study
- Open Access
- 15 July 2024
- Published by American Association for Cancer Research (AACR)
Supplementary Figure S5 from Biomarker Identification and Risk Prediction Model Development for Differentiated Thyroid Carcinoma Lung Metastasis Based on Primary Lesion Proteomics
Figure S5. Protein expression categorized into 4 levels.
- 15 July 2024
- Published by American Association for Cancer Research (AACR)
Open AccessSupplementary Figure S5 from Biomarker Identification and Risk Prediction Model Development for Differentiated Thyroid Carcinoma Lung Metastasis Based on Primary Lesion Proteomics
Figure S5. Protein expression categorized into 4 levels.
- Open Access
- 15 July 2024
- Published by American Association for Cancer Research (AACR)
Supplementary Table S2 from Biomarker Identification and Risk Prediction Model Development for Differentiated Thyroid Carcinoma Lung Metastasis Based on Primary Lesion Proteomics
Table S2. Relevant clinical information of 44 LMDTC and 25 NMDTC patients used for 4D Label-free proteomic analysis, PRM, and IHC dual validation
- 15 July 2024
- Published by American Association for Cancer Research (AACR)
Open AccessSupplementary Table S2 from Biomarker Identification and Risk Prediction Model Development for Differentiated Thyroid Carcinoma Lung Metastasis Based on Primary Lesion Proteomics
Table S2. Relevant clinical information of 44 LMDTC and 25 NMDTC patients used for 4D Label-free proteomic analysis, PRM, and IHC dual validation
- Open Access
- 15 July 2024
- Published by American Association for Cancer Research (AACR)
Supplementary Table S6 from Biomarker Identification and Risk Prediction Model Development for Differentiated Thyroid Carcinoma Lung Metastasis Based on Primary Lesion Proteomics
Table S6. Proteins with degree>40 in the PPI network
- 15 July 2024
- Published by American Association for Cancer Research (AACR)
Open AccessSupplementary Table S6 from Biomarker Identification and Risk Prediction Model Development for Differentiated Thyroid Carcinoma Lung Metastasis Based on Primary Lesion Proteomics
Table S6. Proteins with degree>40 in the PPI network
- Open Access
- 15 July 2024
- Published by American Association for Cancer Research (AACR)
Supplementary Figure S4 from Biomarker Identification and Risk Prediction Model Development for Differentiated Thyroid Carcinoma Lung Metastasis Based on Primary Lesion Proteomics
Figure S4. WGCNA analysis of the proteomics data.
- 15 July 2024
- Published by American Association for Cancer Research (AACR)
Open AccessSupplementary Figure S4 from Biomarker Identification and Risk Prediction Model Development for Differentiated Thyroid Carcinoma Lung Metastasis Based on Primary Lesion Proteomics
Figure S4. WGCNA analysis of the proteomics data.
- Open Access
- 15 July 2024
- Published by American Association for Cancer Research (AACR)
Supplementary Material S1 from Biomarker Identification and Risk Prediction Model Development for Differentiated Thyroid Carcinoma Lung Metastasis Based on Primary Lesion Proteomics
KEGG Analysis Table of Differential Proteins
- 15 July 2024
- Published by American Association for Cancer Research (AACR)
Open AccessSupplementary Material S1 from Biomarker Identification and Risk Prediction Model Development for Differentiated Thyroid Carcinoma Lung Metastasis Based on Primary Lesion Proteomics
KEGG Analysis Table of Differential Proteins
- Open Access
- 15 July 2024
- Published by American Association for Cancer Research (AACR)
Supplementary Table S3 from Biomarker Identification and Risk Prediction Model Development for Differentiated Thyroid Carcinoma Lung Metastasis Based on Primary Lesion Proteomics
Table S3. Conditions for setting up PRM quantitative peptide segments
- 15 July 2024
- Published by American Association for Cancer Research (AACR)
Open AccessSupplementary Table S3 from Biomarker Identification and Risk Prediction Model Development for Differentiated Thyroid Carcinoma Lung Metastasis Based on Primary Lesion Proteomics
Table S3. Conditions for setting up PRM quantitative peptide segments
- Open Access
- 15 July 2024
- Published by American Association for Cancer Research (AACR)
Supplementary Figure S2 from Biomarker Identification and Risk Prediction Model Development for Differentiated Thyroid Carcinoma Lung Metastasis Based on Primary Lesion Proteomics
Figure S2. Principal component analysis of patients in LMDTC (M) group and NMDTC (NM) group.
- 15 July 2024
- Published by American Association for Cancer Research (AACR)
Open AccessSupplementary Figure S2 from Biomarker Identification and Risk Prediction Model Development for Differentiated Thyroid Carcinoma Lung Metastasis Based on Primary Lesion Proteomics
Figure S2. Principal component analysis of patients in LMDTC (M) group and NMDTC (NM) group.