Related Experiment Video
Updated: May 16, 2025
04:23
A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images
Published on: April 21, 2023
1.7K
A predictive nomogram of thyroid nodules based on deep learning ultrasound image analysis
1Department of Ultrasound, the Second Hospital, Cheeloo College of Medicine, Shandong University, Jinan, Shandong, China.
Frontiers in Endocrinology
|May 14, 2025
Summary
A novel deep learning nomogram model accurately distinguishes benign from malignant thyroid nodules using ultrasound characteristics. This model, integrating age, echogenic foci, and deep learning predictions, outperforms existing classification systems.
Area of Science:
- Medical Imaging
- Artificial Intelligence in Medicine
- Oncology
Background:
- Thyroid nodules are common, with accurate differentiation between benign and malignant types crucial for patient management.
- Current diagnostic methods, including ultrasound and classification systems like C-TIRADS, have limitations in predictive performance.
Purpose of the Study:
- To develop and validate a deep learning-based nomogram model for improved prediction of thyroid nodule malignancy.
- To compare the diagnostic accuracy of the proposed nomogram model against existing methods.
Main Methods:
- Retrospective analysis of 2247 thyroid nodules with detailed clinical and ultrasound characteristics.
- Development of deep learning models for ultrasound image analysis.
- Construction of a nomogram model integrating deep learning predictions with significant clinical ultrasound features (age, echogenic foci).
Main Results:
- The deep learning model achieved high accuracy (0.886), precision (0.900), recall (0.889), and F1-score (0.895).
- Age, echogenic foci, and deep learning predicted values were identified as significant independent predictors (P<0.05).
- The nomogram model demonstrated superior accuracy, calibration, and clinical utility compared to C-TIRADS and deep learning models alone.
Conclusions:
- Age, deep learning predicted values, and echogenic foci are effective independent predictors for differentiating thyroid nodules.
- The integrated nomogram model offers enhanced diagnostic accuracy for thyroid nodules, improving upon current standards.
