Deep Learning-Based Intraoperative Dual-tracer Video Analysis of Sentinel Lymph Node Mapping for Metastasis Prediction in cN0 Papillary Thyroid Carcinoma

Deep Learning-Based Intraoperative Dual-tracer Video Analysis of Sentinel Lymph Node Mapping for Metastasis Prediction in cN0 Papillary Thyroid Carcinoma: A Prospective Cohort Study

The goal of this observational study is to learn if a computer program (deep learning) can accurately predict lymph node spread in adults with papillary thyroid cancer who have no signs of lymph node involvement before surgery (called cN0). The main questions it aims to answer are:

• Can video analysis of lymph node mapping during surgery predict if cancer has spread to lymph nodes beyond the first-draining (sentinel) lymph node?
• Can this prediction help surgeons decide how much tissue to remove during surgery?

During surgery, participants will receive an injection of two special dyes (carbon nanoparticles and indocyanine green) near the thyroid tumor. These dyes travel through the lymphatic system and help surgeons see the lymph nodes. A special camera records a video of how the dyes move and light up the lymph nodes.

Researchers will use computer programs to analyze these videos along with other medical information (such as ultrasound results and tumor characteristics) to predict whether cancer has spread to additional lymph nodes. The predictions will be compared against the actual results from tissue samples examined after surgery.

Participants will receive standard thyroid cancer surgery. The study does not change the surgical treatment. The video recording adds no extra risk to participants.

Study Overview

• Status: Completed
• Conditions: Papillary Thyroid Carcinoma
• Intervention / Treatment: Diagnostic test: Carbon nanoparticle (CNs) sentinel lymph node mapping; Diagnostic test: Indocyanine green (ICG) sentinel lymph node mapping; Diagnostic test: Dual-tracer (ICG combined with CNs) sentinel lymph node mapping

Detailed Description

BACKGROUND AND RATIONALE:

Papillary thyroid carcinoma (PTC) is one of the fastest-growing cancers worldwide. A major challenge in treating PTC is that 30% to 80% of patients who appear to have no lymph node involvement before surgery (clinically node-negative, or cN0) actually have hidden (occult) cancer spread to their lymph nodes. Current imaging methods like ultrasound often miss these small areas of cancer spread.

This creates a difficult decision for surgeons: removing too many lymph nodes increases the risk of complications such as damage to the parathyroid glands (which control calcium levels) and the nerves that control the voice. However, removing too few lymph nodes may leave cancer behind, which can lead to recurrence.

Sentinel lymph node (SLN) mapping is a technique that identifies the first lymph nodes that drain from a tumor. The idea is that if cancer spreads through the lymphatic system, it will reach these sentinel nodes first. However, current single-tracer methods for SLN mapping in thyroid cancer have limitations and variable results.

This study uses a dual-tracer approach that combines two different dyes:

1. Carbon nanoparticles (CNs): These provide long-lasting black staining that helps surgeons see lymph nodes clearly
2. Indocyanine green (ICG): This dye glows under near-infrared light, allowing real-time visualization of lymphatic flow

By combining these two tracers, surgeons can see both the structure of lymph nodes and how lymphatic fluid flows through them over time.

STUDY DESIGN:

This is a prospective, single-center, observational cohort study. The study does not change the surgical treatment that participants receive. All participants undergo standard thyroid cancer surgery with lymph node removal as determined by their surgical team.

STUDY PROCEDURES:

1. Pre-operative Assessment:

   All participants undergo standard pre-operative evaluation including:

   • Physical examination
   • Thyroid ultrasound with detailed lymph node assessment
   • Fine-needle aspiration biopsy to confirm PTC diagnosis
   • Genetic testing for common mutations (such as BRAF)
   • Standard blood tests

2. Surgical Procedure:

   During surgery, participants receive the dual-tracer injection under ultrasound guidance. The injection is given at multiple points around the thyroid tumor. The specific preparation is:

   • 0.1 ml of ICG solution (concentration: 2.5 mg/ml)
   • 0.1 ml of carbon nanoparticle suspension (concentration: 50 mg/ml) These are mixed together and injected using precise, multi-point technique.

3. Video Recording:

   A near-infrared fluorescence imaging system records the entire process of lymph node visualization. The recording captures:

   • How the dyes spread through the lymphatic channels
   • When each lymph node first becomes visible
   • How the fluorescence signal changes over time
   • The pattern of lymphatic drainage

   Videos are recorded at high resolution (1920 × 1080 pixels) at approximately 30 frames per second. A standardized 3-minute segment is extracted from each video for analysis, providing 150 frames per patient.

4. Surgical Decisions:

   The sentinel lymph node (the first node that lights up) is removed and sent for immediate frozen section analysis. Based on standard criteria, surgeons decide whether to perform:

   • Ipsilateral central lymph node dissection (removing lymph nodes on the same side as the tumor)
   • Lateral lymph node dissection (if certain criteria are met)
   • Contralateral central dissection (removing lymph nodes on the opposite side)

   These decisions follow the standard surgical protocol at our institution and are not influenced by the deep learning predictions.

5. Pathological Examination:

All removed lymph nodes are examined by pathologists to determine:

• Whether the sentinel lymph node contains cancer (SLNM status)
• Whether second-echelon lymph nodes contain cancer (SeLNM)
• Whether non-sentinel lymph nodes contain cancer (NsLNM)
• The number and location of all positive lymph nodes

DATA COLLECTION AND ANALYSIS:

Clinical Data (32 variables):

• Demographics: age, sex, body mass index
• Ultrasound features: tumor size, location, margins, calcification patterns, aspect ratio, TI-RADS classification
• Pathological features: multifocality, extrathyroidal extension, capsular invasion
• Genetic data: BRAF mutation status and other relevant mutations

Video Analysis:

Two experienced surgeons (each with more than 10 years of experience) manually identify and outline the regions of interest (the sentinel lymph nodes) in each video frame. This creates 19,650 mask images across all participants.

Feature Extraction:

The deep learning system extracts multiple types of features:

Spatial Features (2,048 dimensions):

• Image features from a pre-trained neural network (EfficientNet-B5)
• Grayscale characteristics
• Shape and morphological features
• Hu moment descriptors (mathematical descriptions of shape)

Temporal Features (20 dimensions):

• Frame-to-frame differences showing how the image changes
• Optical flow measurements showing movement patterns
• Fluorescence intensity changes over time
• Flow velocity measurements at specific time points

DEEP LEARNING MODELS:

Nine different deep learning architectures are developed and compared:

1. Convolutional Neural Network (CNN): Extracts local spatial features from images
2. Long Short-Term Memory (LSTM): Captures patterns in time-series data
3. CNN + LSTM: Combines spatial and temporal feature extraction
4. CNN + LSTM + Attention: Adds attention mechanism to focus on important time points
5. Transformer: Uses self-attention to capture global patterns in sequential data
6. Crossformer: A specialized architecture for time-series analysis
7. 3D-CNN: Processes video data as three-dimensional volumes
8. LSTM + Transformer: Hybrid model combining LSTM and Transformer strengths
9. LSTM + Crossformer: Hybrid model combining LSTM and Crossformer

All models use:

• Binary cross-entropy loss function
• Adam optimizer for training
• Data augmentation (rotation, scaling, noise injection) to prevent overfitting
• Weighted loss functions to handle class imbalance
• Dropout regularization

MODEL EVALUATION:

Models are evaluated using 10-fold stratified cross-validation, ensuring balanced distribution of outcomes in training and testing sets. Performance metrics include:

• Area Under the ROC Curve (AUC): Measures overall discrimination ability
• Accuracy: Percentage of correct predictions
• Sensitivity: Ability to correctly identify patients with metastasis
• Specificity: Ability to correctly identify patients without metastasis
• Positive Predictive Value: Probability that a positive prediction is correct
• Negative Predictive Value: Probability that a negative prediction is correct
• F1 Score: Balance between precision and recall
• Brier Score: Calibration of predicted probabilities

Additional analyses include:

• Decision curve analysis to assess clinical utility
• Calibration curves to check prediction reliability
• Learning curves to assess overfitting
• DeLong test for statistical comparison between models
• Probability-based model ranking approach (PMRA)

MODEL INTERPRETABILITY:

To understand how the model makes predictions, we use SHapley Additive exPlanations (SHAP) analysis. This technique:

• Identifies which features contribute most to predictions
• Shows how each feature affects individual predictions
• Reveals whether the model relies on clinically meaningful factors
• Provides transparency into the "black box" of deep learning

OUTCOMES:

Primary Outcomes:

1. Second-echelon lymph node metastasis (SeLNM): Cancer spread to lymph nodes beyond the sentinel node in the drainage pathway
2. Non-sentinel lymph node metastasis (NsLNM): Cancer spread to any lymph node other than the sentinel node

Both outcomes are determined by final pathological examination of surgically removed tissue (the gold standard).

Secondary Outcomes:

• Model performance metrics (AUC, sensitivity, specificity, etc.)
• Feature importance rankings from SHAP analysis
• Comparison of model architectures

STATISTICAL CONSIDERATIONS:

Sample Size:

Based on power calculations assuming:

• Alpha = 0.05 (significance level)
• Power = 0.80
• Expected sensitivity difference of 20% between methods
• Expected prevalence of lymph node metastasis of 50%

A minimum of 335 participants was calculated. Due to strict inclusion criteria and video quality requirements, 131 participants with complete, high-quality data were included in the final analysis.

Statistical Methods:

• Descriptive statistics for baseline characteristics
• Chi-square or Fisher exact test for categorical variables
• Mann-Whitney U test for continuous variables
• DeLong test for comparing AUC values between models
• 95% confidence intervals for all performance metrics

FOLLOW-UP:

While the primary analysis focuses on intraoperative prediction, participants are followed according to standard clinical care protocols. Long-term outcomes including recurrence-free survival may be analyzed in future studies.

ETHICAL CONSIDERATIONS:

This study was approved by the Ethics Committee of the First Affiliated Hospital of Chongqing Medical University (Approval No. 2023-322). All participants provided written informed consent before enrollment.

The study poses minimal additional risk to participants because:

• The dual-tracer injection uses agents already approved for clinical use
• The video recording is non-invasive and does not affect surgical decisions
• All surgical decisions are made according to standard protocols
• The deep learning analysis is performed retrospectively and does not influence treatment

POTENTIAL IMPACT:

If successful, this approach could:

1. Help surgeons make more informed decisions about the extent of lymph node removal
2. Reduce unnecessary extensive surgery in low-risk patients
3. Ensure adequate surgery in high-risk patients who might otherwise be undertreated
4. Provide real-time, objective decision support during surgery
5. Standardize the interpretation of lymphatic mapping across different surgeons and centers
6. Serve as a training tool for less experienced surgeons

LIMITATIONS:

• Single-center study, which may limit generalizability
• Relatively small sample size due to strict quality requirements
• Manual annotation of regions of interest may introduce variability
• Deep learning models require validation in external cohorts
• Real-time implementation during surgery requires further development

FUTURE DIRECTIONS:

• External validation in multiple centers
• Development of real-time prediction software for intraoperative use
• Integration with other imaging modalities
• Prospective interventional trials to assess clinical benefit of AI-guided surgery

• Study Type: Observational
• Enrollment (Actual): 131

Contacts and Locations

Study Locations

• China
  • Chongqing, China, 400016
    • The First Affiliated Hospital of Chongqing Medical University

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: Adult; Older Adult
• Accepts Healthy Volunteers: No

• Sampling Method: Non-Probability Sample

Study Population

Adults diagnosed with clinically node-negative papillary thyroid carcinoma (cN0-PTC) who are scheduled to undergo surgical treatment at the Department of Breast and Thyroid Surgery, First Affiliated Hospital of Chongqing Medical University, Chongqing, China. Participants are identified through routine clinical care and preoperative evaluation. The study population includes patients with primary PTC confirmed by fine-needle aspiration biopsy who have no evidence of lymph node metastasis on preoperative ultrasound or other imaging studies. Both male and female adults aged 18 years and older are eligible regardless of tumor size, location, or genetic mutation status.

Description

Inclusion Criteria:

1. Age 18 years or older at the time of enrollment
2. Histologically confirmed papillary thyroid carcinoma (PTC) by preoperative fine-needle aspiration biopsy
3. Clinically node-negative (cN0) status confirmed by preoperative imaging (ultrasound and/or cross-sectional imaging showing no evidence of lymph node metastasis)
4. Scheduled to undergo thyroid surgery with simultaneous central lymph node dissection
5. Willing and able to provide written informed consent

6. Complete preoperative clinical data available, including:

   • Demographic information (age, sex, body mass index)
   • Thyroid ultrasound report with detailed tumor characteristics
   • Fine-needle aspiration biopsy pathology report
   • Genetic testing results (when available)
7. Able to undergo intraoperative dual-tracer sentinel lymph node mapping with near-infrared fluorescence video recording

Exclusion Criteria:

1. History of previous neck surgery (including thyroid surgery, parathyroid surgery, or other cervical operations)
2. History of external beam radiotherapy to the head and neck region
3. Diagnosis of thyroid malignancy other than papillary thyroid carcinoma (e.g., follicular thyroid carcinoma, medullary thyroid carcinoma, anaplastic thyroid carcinoma, or thyroid lymphoma)
4. Known allergy or hypersensitivity to indocyanine green (ICG), iodine, or carbon nanoparticles
5. Severe hepatic insufficiency (ICG is metabolized by the liver)
6. Pregnancy or breastfeeding
7. Incomplete medical records or missing essential preoperative data
8. Refusal to undergo dual-tracer imaging procedure
9. Inability to obtain satisfactory intraoperative near-infrared fluorescence video recording due to technical issues or poor image quality
10. Participation in another interventional clinical trial that may interfere with the current study
11. Any condition that, in the investigator's opinion, would compromise the participant's safety or the quality of the study data

Study Plan

How is the study designed?

Number of groups / cohorts

3

Cohorts and Interventions

Group / Cohort	Intervention / Treatment
ICG Group; Adults with clinically node-negative papillary thyroid carcinoma (cN0-PTC) who undergo intraoperative sentinel lymph node mapping using indocyanine green (ICG) alone. Intervention: 0.2 ml of ICG solution (concentration: 2.5 mg/ml) injected at multiple points around the thyroid tumor under ultrasound guidance. Near-infrared fluorescence imaging is used to visualize lymphatic drainage and identify sentinel lymph nodes. Participants undergo standard thyroid surgery including thyroid lobectomy and central lymph node dissection, with additional dissection based on intraoperative findings.	Diagnostic test: Carbon nanoparticle (CNs) sentinel lymph node mapping; Intraoperative sentinel lymph node mapping using carbon nanoparticle suspension with visual identification. Preparation: Carbon nanoparticle suspension is used at the commercial concentration of 50 mg/ml. Administration: 0.2 ml of carbon nanoparticle suspension is injected at multiple points around the thyroid tumor under real-time ultrasound guidance using a precision multi-point stereotactic injection technique. Visualization: Carbon nanoparticles (diameter 150 nm) selectively enter lymphatic channels and accumulate in lymph nodes, producing visible black staining. Surgeons identify sentinel lymph nodes by direct visual inspection of black-stained nodes. The sentinel lymph node is defined as the first lymph node that shows black staining after tracer injection.; Other Names: Carbon nanoparticle suspension lymph node tracing; Nano-carbon lymphatic mapping
CNs Group; Adults with clinically node-negative papillary thyroid carcinoma (cN0-PTC) who undergo intraoperative sentinel lymph node mapping using carbon nanoparticles (CNs) alone. Intervention: 0.2 ml of carbon nanoparticle suspension (concentration: 50 mg/ml) injected at multiple points around the thyroid tumor under ultrasound guidance. Black staining is used to visualize lymph nodes during surgery. Participants undergo standard thyroid surgery including thyroid lobectomy and central lymph node dissection, with additional dissection based on intraoperative findings.	Diagnostic test: Indocyanine green (ICG) sentinel lymph node mapping; Intraoperative sentinel lymph node mapping using indocyanine green (ICG) with near-infrared fluorescence imaging. Preparation: ICG powder (25 mg) is dissolved in 10 ml sterile water to achieve a concentration of 2.5 mg/ml. Administration: 0.2 ml of ICG solution is injected at multiple points around the thyroid tumor under real-time ultrasound guidance using a precision multi-point stereotactic injection technique. Visualization: A near-infrared fluorescence imaging system (excitation wavelength 750-800 nm, emission wavelength 820 nm) is used to visualize lymphatic channels and identify sentinel lymph nodes in real time during surgery. The sentinel lymph node is defined as the first lymph node that shows fluorescence signal after tracer injection.; Other Names: ICG fluorescence lymphography; Near-infrared fluorescence sentinel lymph node identification
ICG+CNs Group (Dual-Tracer); Adults with clinically node-negative papillary thyroid carcinoma (cN0-PTC) who undergo intraoperative sentinel lymph node mapping using combined indocyanine green and carbon nanoparticles (dual-tracer technique). Intervention: A mixture of 0.1 ml ICG solution (concentration: 2.5 mg/ml) and 0.1 ml carbon nanoparticle suspension (concentration: 50 mg/ml) injected at multiple points around the thyroid tumor under ultrasound guidance. Near-infrared fluorescence imaging combined with visual black staining is used to visualize lymphatic drainage and identify sentinel lymph nodes. This group also undergoes deep learning video analysis of the fluorescence imaging process to predict lymph node metastasis. Participants undergo standard thyroid surgery including thyroid lobectomy and central lymph node dissection, with additional dissection based on intraoperative findings.	Diagnostic test: Dual-tracer (ICG combined with CNs) sentinel lymph node mapping; Intraoperative sentinel lymph node mapping using combined indocyanine green and carbon nanoparticles with near-infrared fluorescence imaging and visual identification. Preparation: 0.1 ml of ICG solution (2.5 mg/ml) is mixed with 0.1 ml of carbon nanoparticle suspension (50 mg/ml) to form a 0.2 ml dual-tracer composite agent. Administration: The mixed tracer is injected at multiple points around the thyroid tumor under real-time ultrasound guidance using a precision multi-point stereotactic injection technique. Visualization: Near-infrared fluorescence imaging captures real-time lymphatic flow dynamics (ICG component), while black staining provides durable visual lymph node identification (CNs component). Video recording documents the entire sentinel lymph node visualization process for at least 5 minutes at 1920x1080 resolution. Deep learning analysis: In this group, video recordings are analyzed using nine deep learning models to extract spatiotemporal features and predict second; Other Names: Combined ICG and carbon nanoparticle lymphatic mapping; Dual-tracer fluorescence and visual lymph node identification

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Sentinel Lymph Node Metastasis (SLNM) Status	The presence or absence of cancer metastasis in the sentinel lymph node, determined by postoperative histopathological examination (paraffin section) as the gold standard. SLNM is classified as positive (macrometastasis or micrometastasis present) or negative (no metastasis). The SLNM rate is calculated as: number of participants with positive sentinel lymph nodes divided by total number of participants with successfully identified sentinel lymph nodes.	immediately after the surgery
Sentinel Lymph Node Detection Rate	The proportion of participants in whom sentinel lymph nodes are successfully identified using each tracer method (ICG alone, CNs alone, or ICG+CNs dual-tracer). A sentinel lymph node is defined as the first lymph node visualized after tracer injection. Detection rate is calculated as: number of participants with successfully identified sentinel lymph nodes divided by total number of participants in each group, expressed as a percentage.	immediately after the surgery
Second-Echelon Lymph Node Metastasis (SeLNM)	The presence or absence of cancer metastasis in second-echelon lymph nodes (lymph nodes beyond the sentinel node in the lymphatic drainage pathway), determined by postoperative histopathological examination. SeLNM is the primary prediction target for the deep learning models in the ICG+CNs group. SeLNM status is classified as positive or negative based on paraffin section pathology results.	perioperatively
Non-Sentinel Lymph Node Metastasis (NsLNM)	The presence or absence of cancer metastasis in any lymph node other than the sentinel lymph node, determined by postoperative histopathological examination. NsLNM includes metastasis in central compartment nodes (prelaryngeal, pretracheal, paratracheal, and nodes posterior to recurrent laryngeal nerve) and lateral compartment nodes when dissected. NsLNM is the second primary prediction target for the deep learning models in the ICG+CNs group.	perioperatively

Secondary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Sensitivity of Sentinel Lymph Node Mapping	The ability of each tracer method to correctly identify participants who have lymph node metastasis. Sensitivity is calculated as: true positives divided by (true positives + false negatives), expressed as a percentage. A true positive is defined as a positive sentinel lymph node in a participant with confirmed central lymph node metastasis on final pathology. Compared among ICG, CNs, and ICG+CNs groups.	through study completion, an average of 1 year
Specificity of Sentinel Lymph Node Mapping	The ability of each tracer method to correctly identify participants who do not have lymph node metastasis. Specificity is calculated as: true negatives divided by (true negatives + false positives), expressed as a percentage. A true negative is defined as a negative sentinel lymph node in a participant with no central lymph node metastasis on final pathology. Compared among ICG, CNs, and ICG+CNs groups.	through study completion, an average of 1 year
Positive Predictive Value (PPV) of Sentinel Lymph Node Mapping	The probability that participants with a positive sentinel lymph node truly have central lymph node metastasis. PPV is calculated as: true positives divided by (true positives + false positives), expressed as a percentage. Compared among ICG, CNs, and ICG+CNs groups.	through study completion, an average of 1 year
Negative Predictive Value (NPV) of Sentinel Lymph Node Mapping	The probability that participants with a negative sentinel lymph node truly do not have central lymph node metastasis. NPV is calculated as: true negatives divided by (true negatives + false negatives), expressed as a percentage. A high NPV indicates that a negative sentinel lymph node reliably rules out metastatic disease. Compared among ICG, CNs, and ICG+CNs groups.	through study completion, an average of 1 year
Deep Learning Model Performance - Area Under ROC Curve (AUC)	The area under the receiver operating characteristic curve for each deep learning model (CNN, LSTM, CNN+LSTM, CNN+LSTM+Attention, Transformer, Crossformer, 3D-CNN, LSTM+Transformer, LSTM+Crossformer) in predicting SeLNM and NsLNM in the ICG+CNs group. AUC ranges from 0 to 1, with higher values indicating better discrimination. Evaluated using 10-fold stratified cross-validation.	through study completion, an average of 1 year
Deep Learning Model Performance - Accuracy, Sensitivity, and Specificity	Performance metrics of the optimal deep learning model for predicting SeLNM and NsLNM in the ICG+CNs group. Accuracy is the proportion of correct predictions. Sensitivity is the proportion of actual positive cases correctly identified. Specificity is the proportion of actual negative cases correctly identified. All metrics expressed as percentages with 95% confidence intervals.	through study completion, an average of 1 year

Other Outcome Measures

Outcome Measure	Measure Description	Time Frame
Number of Sentinel Lymph Nodes Identified	The mean number of sentinel lymph nodes identified per participant using each tracer method. Compared among ICG, CNs, and ICG+CNs groups using appropriate statistical tests.	immediately after surgery
Total Number of Lymph Nodes Retrieved	The total number of lymph nodes retrieved during central and lateral lymph node dissection per participant. Reported as mean with standard deviation for each group.	perioperatively
Feature Importance from SHAP Analysis	Identification and ranking of the most important predictive features contributing to the deep learning model predictions, determined by SHapley Additive exPlanations (SHAP) analysis. Features include temporal fluorescence-flow characteristics, spatial structural features, and clinical variables. Reported as mean absolute SHAP values for top contributing features.	through study completion, an average of 1 year
Surgical Complications	Incidence of surgery-related complications including: transient or permanent hypoparathyroidism (based on postoperative calcium and parathyroid hormone levels), transient or permanent recurrent laryngeal nerve injury (based on postoperative laryngoscopy), postoperative bleeding requiring intervention, and wound infection. Reported as number and percentage of participants in each group.	up to 24 weeks
Tracer-Related Adverse Events	Incidence of adverse events related to tracer injection, including allergic reactions, injection site reactions, and any other tracer-associated complications. ICG-related adverse events are expected to be less than 0.05% based on published literature. Reported as number and percentage of participants in each group.	perioperatively

Collaborators and Investigators

• Sponsor: First Affiliated Hospital of Chongqing Medical University
• Investigators: Study Director: Xinliang Su, MD,PhD, First Affiliated Hospital of Chongqing Medical University; Study Chair: Han Gao, MD,PhD, Women and Children's Hospital of Chongqing Medical University; Principal Investigator: Xinliang Su, MD,PhD, First Affiliated Hospital of Chongqing Medical University

Publications and helpful links

General Publications

• Qian T, Zhou Y, Yao J, Ni C, Asif S, Chen C, Lv L, Ou D, Xu D. Deep learning based analysis of dynamic video ultrasonography for predicting cervical lymph node metastasis in papillary thyroid carcinoma. Endocrine. 2025 Mar;87(3):1060-1069. doi: 10.1007/s12020-024-04091-w. Epub 2024 Nov 18.
• Ding X, Liu Y, Zhao J, Wang R, Li C, Luo Q, Shen C. A novel wavelet-transform-based convolution classification network for cervical lymph node metastasis of papillary thyroid carcinoma in ultrasound images. Comput Med Imaging Graph. 2023 Oct;109:102298. doi: 10.1016/j.compmedimag.2023.102298. Epub 2023 Sep 9.
• Yang D, Li T, Li L, Chen S, Li X. Multi-modal convolutional neural network-based thyroid cytology classification and diagnosis. Hum Pathol. 2025 Jul;161:105868. doi: 10.1016/j.humpath.2025.105868. Epub 2025 Jul 4.
• Chu X, Wang T, Chen M, Li J, Wang L, Wang C, Wang H, Wong ST, Chen Y, Li H. Deep learning model for malignancy prediction of TI-RADS 4 thyroid nodules with high-risk characteristics using multimodal ultrasound: A multicentre study. Comput Med Imaging Graph. 2025 Sep;124:102576. doi: 10.1016/j.compmedimag.2025.102576. Epub 2025 May 26.
• Liang M, Zhu T, Huang N, Zhang L, Yang C, Gao H, Zhang X, Li P, Cheng M, Wang K. Incorporating sentinel chain involvement pattern to predict non-sentinel lymph nodes status in breast cancer after neoadjuvant chemotherapy. Clin Transl Oncol. 2026 Jan;28(1):203-214. doi: 10.1007/s12094-025-03993-z. Epub 2025 Jul 31.
• Mittendorf EA, Hunt KK, Boughey JC, Bassett R, Degnim AC, Harrell R, Yi M, Meric-Bernstam F, Ross MI, Babiera GV, Kuerer HM, Hwang RF. Incorporation of sentinel lymph node metastasis size into a nomogram predicting nonsentinel lymph node involvement in breast cancer patients with a positive sentinel lymph node. Ann Surg. 2012 Jan;255(1):109-15. doi: 10.1097/SLA.0b013e318238f461.
• Zhou L, Yao J, Ou D, Li M, Lei Z, Wang L, Xu D. A multi-institutional study of association of sonographic characteristics with cervical lymph node metastasis in unifocal papillary thyroid carcinoma. Front Endocrinol (Lausanne). 2022 Sep 23;13:965241. doi: 10.3389/fendo.2022.965241. eCollection 2022.
• Khan SU, Fatima K, Malik F, Kalkavan H, Wani A. Cancer metastasis: Molecular mechanisms and clinical perspectives. Pharmacol Ther. 2023 Oct;250:108522. doi: 10.1016/j.pharmthera.2023.108522. Epub 2023 Sep 1.
• Cao R, Ji H, Feng N, Zhang Y, Yang X, Andersson P, Sun Y, Tritsaris K, Hansen AJ, Dissing S, Cao Y. Collaborative interplay between FGF-2 and VEGF-C promotes lymphangiogenesis and metastasis. Proc Natl Acad Sci U S A. 2012 Sep 25;109(39):15894-9. doi: 10.1073/pnas.1208324109. Epub 2012 Sep 11.
• Li Y, Chen H, Zhao Y, Yan Q, Chen L, Song Q. circUBE2G1 interacts with hnRNPU to promote VEGF-C-mediated lymph node metastasis of lung adenocarcinoma. Front Oncol. 2024 Nov 27;14:1455909. doi: 10.3389/fonc.2024.1455909. eCollection 2024.
• Guang Y, Wan F, He W, Zhang W, Gan C, Dong P, Zhang H, Zhang Y. A model for predicting lymph node metastasis of thyroid carcinoma: a multimodality convolutional neural network study. Quant Imaging Med Surg. 2023 Dec 1;13(12):8370-8382. doi: 10.21037/qims-23-318. Epub 2023 Nov 7.
• Luo QW, Gao S, Lv X, Li SJ, Wang BF, Han QQ, Wang YP, Guan QL, Gong T. A novel tool for predicting the risk of central lymph node metastasis in patients with papillary thyroid microcarcinoma: a retrospective cohort study. BMC Cancer. 2022 Jun 2;22(1):606. doi: 10.1186/s12885-022-09655-5.
• Popadich A, Levin O, Lee JC, Smooke-Praw S, Ro K, Fazel M, Arora A, Tolley NS, Palazzo F, Learoyd DL, Sidhu S, Delbridge L, Sywak M, Yeh MW. A multicenter cohort study of total thyroidectomy and routine central lymph node dissection for cN0 papillary thyroid cancer. Surgery. 2011 Dec;150(6):1048-57. doi: 10.1016/j.surg.2011.09.003.
• Huang C, Cong S, Shang S, Wang M, Zheng H, Wu S, An X, Liang Z, Zhang B. Web-Based Ultrasonic Nomogram Predicts Preoperative Central Lymph Node Metastasis of cN0 Papillary Thyroid Microcarcinoma. Front Endocrinol (Lausanne). 2021 Sep 7;12:734900. doi: 10.3389/fendo.2021.734900. eCollection 2021.
• Wang Y, Deng C, Shu X, Yu P, Wang H, Su X, Tan J. Risk Factors and a Prediction Model of Lateral Lymph Node Metastasis in CN0 Papillary Thyroid Carcinoma Patients With 1-2 Central Lymph Node Metastases. Front Endocrinol (Lausanne). 2021 Oct 15;12:716728. doi: 10.3389/fendo.2021.716728. eCollection 2021.
• Chun L, Wang D, He L, Li D, Fu Z, Xue S, Su X, Zhou J. Explainable machine learning model for predicting paratracheal lymph node metastasis in cN0 papillary thyroid cancer. Sci Rep. 2024 Sep 27;14(1):22361. doi: 10.1038/s41598-024-73837-3.
• Du W, Fang Q, Zhang X, Dai L. Metastasis of cN0 Papillary Thyroid Carcinoma of the Isthmus to the Lymph Node Posterior to the Right Recurrent Laryngeal Nerve. Front Endocrinol (Lausanne). 2021 May 10;12:677986. doi: 10.3389/fendo.2021.677986. eCollection 2021.
• He F, Chen S, Liu X, Yang X, Qin X. Multimodal Deep Learning Model Based on Ultrasound and Cytological Images Predicts Risk Stratification of cN0 Papillary Thyroid Carcinoma. Acad Radiol. 2025 Sep;32(9):5091-5099. doi: 10.1016/j.acra.2025.06.043. Epub 2025 Jul 14.
• Yin SM, Lien JJ, Chiu IM. Deep learning implementation for extrahepatic bile duct detection during indocyanine green fluorescence-guided laparoscopic cholecystectomy: pilot study. BJS Open. 2025 Mar 4;9(2):zraf013. doi: 10.1093/bjsopen/zraf013.
• Shen B, Zhang Z, Shi X, Cao C, Zhang Z, Hu Z, Ji N, Tian J. Real-time intraoperative glioma diagnosis using fluorescence imaging and deep convolutional neural networks. Eur J Nucl Med Mol Imaging. 2021 Oct;48(11):3482-3492. doi: 10.1007/s00259-021-05326-y. Epub 2021 Apr 27.
• Chang YY, Yang HP, Chen YY, Yen HH. Comparison of the performance between an AI-based vision transformer and human endoscopists in predicting the endoscopic and histologic activities of ulcerative colitis. Digit Health. 2026 Jan 5;12:20552076251412694. doi: 10.1177/20552076251412694. eCollection 2026 Jan-Dec.
• Kumar R, Sethia K, Kumar V. Video-endoscopic versus open inguinal lymphadenectomy: Long-term oncological outcomes in penile cancer. BJUI Compass. 2026 Jan 6;7(1):e70153. doi: 10.1002/bco2.70153. eCollection 2026 Jan.
• Oh N, Kim B, Kim T, Rhu J, Kim J, Choi GS. Real-time segmentation of biliary structure in pure laparoscopic donor hepatectomy. Sci Rep. 2024 Sep 28;14(1):22508. doi: 10.1038/s41598-024-73434-4.
• Jha D, Ali S, Tomar NK, Johansen HD, Johansen D, Rittscher J, Riegler MA, Halvorsen P. Real-Time Polyp Detection, Localization and Segmentation in Colonoscopy Using Deep Learning. IEEE Access. 2021 Mar 4;9:40496-40510. doi: 10.1109/ACCESS.2021.3063716. eCollection 2021.
• Dolidze DD, Shabunin AV, Mumladze RB, Vardanyan AV, Covantsev SD, Shulutko AM, Semikov VI, Isaev KM, Kazaryan AM. A Narrative Review of Preventive Central Lymph Node Dissection in Patients With Papillary Thyroid Cancer - A Necessity or an Excess. Front Oncol. 2022 Jun 29;12:906695. doi: 10.3389/fonc.2022.906695. eCollection 2022.
• Giordano D, Valcavi R, Thompson GB, Pedroni C, Renna L, Gradoni P, Barbieri V. Complications of central neck dissection in patients with papillary thyroid carcinoma: results of a study on 1087 patients and review of the literature. Thyroid. 2012 Sep;22(9):911-7. doi: 10.1089/thy.2012.0011. Epub 2012 Jul 24.
• Zhang L, Cheng M, Lin Y, Zhang J, Shen B, Chen Y, Yang C, Yang M, Zhu T, Gao H, Ji F, Li J, Wang K. Ultrasound-assisted carbon nanoparticle suspension mapping versus dual tracer-guided sentinel lymph node biopsy in patients with early breast cancer (ultraCars): phase III randomized clinical trial. Br J Surg. 2022 Nov 22;109(12):1232-1238. doi: 10.1093/bjs/znac311.
• Bargon CA, Huibers A, Young-Afat DA, Jansen BAM, Borel-Rinkes IHM, Lavalaye J, van Slooten HJ, Verkooijen HM, van Swol CFP, Doeksen A. Sentinel Lymph Node Mapping in Breast Cancer Patients Through Fluorescent Imaging Using Indocyanine Green: The INFLUENCE Trial. Ann Surg. 2022 Nov 1;276(5):913-920. doi: 10.1097/SLA.0000000000005633. Epub 2022 Jul 27.
• Solis O, Addae J, Sweeting R, Meszoely I, Grau A, Kauffmann R, Kelley M, McCaffrey R, Hewitt K. Cost containment analysis of superparamagnetic iron oxide (SPIO) injection in patients with ductal carcinoma in situ. Breast Cancer Res Treat. 2024 Dec;208(3):565-568. doi: 10.1007/s10549-024-07451-2. Epub 2024 Aug 1.
• Madadi M, Khoee S. Magnetite-based Janus nanoparticles, their synthesis and biomedical applications. Wiley Interdiscip Rev Nanomed Nanobiotechnol. 2023 Nov-Dec;15(6):e1908. doi: 10.1002/wnan.1908. Epub 2023 Jun 4.
• Marengo M, Martin CJ, Rubow S, Sera T, Amador Z, Torres L. Radiation Safety and Accidental Radiation Exposures in Nuclear Medicine. Semin Nucl Med. 2022 Mar;52(2):94-113. doi: 10.1053/j.semnuclmed.2021.11.006. Epub 2021 Dec 13.
• Garau LM, Rubello D, Ferretti A, Boni G, Volterrani D, Manca G. Sentinel lymph node biopsy in small papillary thyroid cancer. A review on novel surgical techniques. Endocrine. 2018 Nov;62(2):340-350. doi: 10.1007/s12020-018-1658-5. Epub 2018 Jul 2.
• Garau LM, Rubello D, Morganti R, Boni G, Volterrani D, Colletti PM, Manca G. Sentinel Lymph Node Biopsy in Small Papillary Thyroid Cancer: A Meta-analysis. Clin Nucl Med. 2019 Feb;44(2):107-118. doi: 10.1097/RLU.0000000000002378.
• Santrac N, Markovic I, Medic Milijic N, Goran M, Buta M, Djurisic I, Dzodic R. Sentinel lymph node biopsy in medullary thyroid microcarcinomas. Endocr J. 2020 Mar 28;67(3):295-304. doi: 10.1507/endocrj.EJ19-0409. Epub 2019 Dec 3.
• Boschin IM, Bertazza L, Scaroni C, Mian C, Pelizzo MR. Sentinel lymph node mapping: current applications and future perspectives in thyroid carcinoma. Front Med (Lausanne). 2023 Oct 24;10:1231566. doi: 10.3389/fmed.2023.1231566. eCollection 2023.
• Likhterov I, Reis LL, Urken ML. Central compartment management in patients with papillary thyroid cancer presenting with metastatic disease to the lateral neck: Anatomic pathways of lymphatic spread. Head Neck. 2017 May;39(5):853-859. doi: 10.1002/hed.24568. Epub 2017 Mar 2.
• Lopez-Martinez R. [Immunological characteristics in candidiasis]. Prensa Med Mex. 1974 Nov-Dec;39(11-12):468-71. No abstract available. Spanish.
• Flagiello D, Gibaud A, Dutrillaux B, Poupon MF, Malfoy B. Distinct patterns of all-trans retinoic acid dependent expression of HOXB and HOXC homeogenes in human embryonal and small-cell lung carcinoma cell lines. FEBS Lett. 1997 Oct 6;415(3):263-7. doi: 10.1016/s0014-5793(97)01118-6.
• Galili U. Evolution of alpha 1,3galactosyltransferase and of the alpha-Gal epitope. Subcell Biochem. 1999;32:1-23. doi: 10.1007/978-1-4615-4771-6_1. No abstract available.
• Dogan H, Tasman F, Cehreli ZC. Effect of gutta-percha solvents at different temperatures on the calcium, phosphorus and magnesium levels of human root dentin. J Oral Rehabil. 2001 Aug;28(8):792-6. doi: 10.1046/j.1365-2842.2001.00714.x.
• Izumi AK, Dawson KL. Zabon phytophotodermatitis: first case reports due to Citrus maxima. J Am Acad Dermatol. 2002 May;46(5 Suppl):S146-7. doi: 10.1067/mjd.2002.107489.
• Chapchal G. [On the conservative therapy of joint contractures in hemophilia]. Schweiz Med Wochenschr. 1967 Jun 17;97(24):788-90. No abstract available. German.
• Hardaway BW, Hoffer FA, Rao BN. Needle localization of small pediatric tumors for surgical biopsy. Pediatr Radiol. 2000 May;30(5):318-22. doi: 10.1007/s002470050748.
• Sokoloff MH, Chung LW. Targeting angiogenic pathways involving tumor-stromal interaction to treat advanced human prostate cancer. Cancer Metastasis Rev. 1998-1999;17(4):307-15. doi: 10.1023/a:1006170612253.
• Roller RJ, Herold BC. Characterization of a BHK(TK-) cell clone resistant to postattachment entry by herpes simplex virus types 1 and 2. J Virol. 1997 Aug;71(8):5805-13. doi: 10.1128/JVI.71.8.5805-5813.1997.
• Hansma HG, Clegg DO, Kokkoli E, Oroudjev E, Tirrell M. Analysis of matrix dynamics by atomic force microscopy. Methods Cell Biol. 2002;69:163-93. doi: 10.1016/s0091-679x(02)69012-0. No abstract available.

Study record dates

Study Major Dates

• Study Start (Actual): April 1, 2024
• Primary Completion (Actual): October 31, 2024
• Study Completion (Actual): October 31, 2024

Study Registration Dates

• First Submitted: January 19, 2026
• First Submitted That Met QC Criteria: February 2, 2026
• First Posted (Actual): February 6, 2026

Study Record Updates

• Last Update Posted (Actual): February 6, 2026
• Last Update Submitted That Met QC Criteria: February 2, 2026
• Last Verified: February 1, 2026

More Information

Terms related to this study

• Keywords: Indocyanine green; Sentinel lymph node biopsy; Deep learning; Papillary thyroid carcinoma; Dual-tracer imaging; Carbon nanoparticles
• Additional Relevant MeSH Terms: Endocrine System Diseases; Neoplasms by Site; Neoplasms; Neoplasms by Histologic Type; Thyroid Neoplasms; Endocrine Gland Neoplasms; Head and Neck Neoplasms; Neoplasms, Glandular and Epithelial; Adenocarcinoma; Carcinoma; Thyroid Diseases; Adenocarcinoma, Papillary; Thyroid Cancer, Papillary
• Other Study ID Numbers: 1stChongqingMU-sxl

Plan for Individual participant data (IPD)

• Plan to Share Individual Participant Data (IPD)?: UNDECIDED

IPD Plan Description

A decision regarding individual participant data sharing has not yet been finalized. The study team is currently evaluating:

1. Technical feasibility of de-identifying video-based imaging data
2. Institutional and ethical requirements for data sharing
3. Development of appropriate data use agreements

The IPD sharing plan will be updated prior to publication of the primary study results. Interested researchers may contact the Principal Investigator (suxinliang@21cn.com) to discuss potential future data access.

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: No