Related Experiment Video
Updated: Sep 10, 2025
Tracking the Mammary Architectural Features and Detecting Breast Cancer with Magnetic Resonance Diffusion Tensor Imaging
Published on: December 15, 2014
Differentiating Benign Thyroid Nodules and Papillary Thyroid Carcinoma Using Time-Dependent Diffusion MRI: A
Huanhuan Ren1, Diwei Shi2,3, Junhao Huang1
1Department of Radiology, Chongqing University Cancer Hospital, Chongqing, China.
Background:
Differentiating benign and malignant thyroid nodules is important for treatment planning and prognostic, yet an ideal method is lacking.
Purpose:
To investigate whether microstructural parameters from time-dependent diffusion MRI (td-dMRI) can accurately distinguish between benign and malignant thyroid nodules.
Study Type:
Single-center, prospective.
Population:
232 participants (median age, 48.0 years; IQR, 34.0-53.0) with pathologically diagnosed thyroid nodules.
Field Strength/Sequence:
3.0 T td-dMRI with two oscillating gradient spin-echo sequences and one pulsed gradient spin-echo sequence.
Assessment:
Histopathology was the reference standard for benign and malignant nodules identification. Clinical factors (sex, age, lesion location, and tumor size) and td-dMRI-derived parameters (cell diameter, cellularity, intracellular volume fraction, extracellular diffusivity, and ADCs) were analyzed via univariate and multivariate regression. Chinese Thyroid Imaging Reporting and Data System (TI-RADS) model and two combined models integrating significant clinical and td-dMRI features with and without TI-RADS were developed and compared.
Statistical Tests:
Model performance was evaluated using the area under the receiver operating characteristic curve (AUC). The DeLong test was used to compare model AUCs. Net reclassification improvement (NRI) and integrated discrimination improvement (IDI) tests were employed for performance comparison. A p value < 0.05 was considered significant.
Results:
Tumor size (maximum diameter) (odds ratio [OR], 0.309) and cellularity (OR, 2.430) from td-dMRI were independently associated with malignant thyroid nodules. The AUC of the combined model based on the two features was not significantly different to that of TI-RADS alone (0.847 vs. 0.891, p = 0.169). The combined model integrating tumor maximum diameter, cellularity, and TI-RADS significantly improved diagnostic accuracy compared to TI-RADS alone (AUC: 0.941 vs. 0.891; IDI = 0.134 [95% confidence interval (CI), 0.085-0.183]; NRI = 0.762 [95% CI, 0.522-0.979]).
Data Conclusion:
The combined model integrating tumor maximum diameter, cellularity based on td-dMRI, and TI-RADS has the potential to differentiate between benign thyroid nodules and PTC.
Evidence Level:
1.
Technical Efficacy:
Stage 2.
