Related Experiment Video
Updated: Jun 5, 2025
03:55
Computer-Aided Three-Dimensional Visualization in the Treatment of Locally Advanced Thyroid Cancer
Published on: June 9, 2023
480
Multimodal MRI Deep Learning for Predicting Central Lymph Node Metastasis in Papillary Thyroid Cancer.
Xiuyu Wang1,2, Heng Zhang2, Hang Fan3
1Nurturing Center of Jiangsu Province for State Laboratory of AI Imaging & Interventional Radiology, Department of Radiology, Zhongda Hospital, Medical School, Southeast University, Nanjing 210018, China.
Cancers
|December 17, 2024
Summary
A new deep learning model accurately predicts central lymph node metastasis in papillary thyroid cancer using MRI and patient data. This approach aids surgical decisions and reduces unnecessary treatments for papillary thyroid cancer patients.
Area of Science:
- Oncology
- Radiology
- Medical Imaging
- Artificial Intelligence
Background:
- Central lymph node metastasis (CLNM) is a critical factor in surgical planning for papillary thyroid cancer (PTC).
- Accurate prediction of CLNM is essential to optimize treatment strategies and avoid overtreatment in PTC patients.
Purpose of the Study:
- To develop a predictive model for CLNM in PTC using magnetic resonance imaging (MRI) and clinicopathological data.
- To compare the performance of deep learning (DL) models against traditional machine learning (ML) models for CLNM prediction.
Main Methods:
- Retrospective analysis of preoperative MRI data from 105 PTC patients.
- Development of a deep learning model (AMMCNet) incorporating convolutional neural networks (CNNs) and fusion of MRI images with clinicopathological data.
- Comparison with traditional ML models including support vector machine (SVM), logistic regression (LR), and random forest (RF).
- Evaluation of predictive performance using receiver operator characteristic (ROC) curve analysis and clinical utility via decision curve analysis (DCA).
Main Results:
- Lesion diameter was identified as an independent risk factor for CLNM.
- Among ML models, the random forest (RF) model achieved the highest area under the curve (AUC) of 0.863.
- The proposed DL fusion model demonstrated superior performance with an AUC of 0.891, outperforming all ML fusion models.
Conclusions:
- A fusion model utilizing the AMMCNet architecture, integrating MRI images and clinicopathological data, effectively predicts CLNM in PTC.
- The developed DL model shows significant potential for improving diagnostic accuracy and guiding surgical decisions in PTC management.
