Objective:To evaluate the safety and efficacy of radiofrequency ablation（RFA）in the treatment of papillary thyroid microcarcinoma（PTMC）near and remote from the thyroid capsule.Methods:The clinical data，ultrasound images，ablation parameters and postoperative follow-up metrics were prospectively collected from 130 patients with pathologically confirmed PTMC in the Second Affiliated Hospital of Xi'an Jiaotong University who underwent RFA treatment between January 2023 and December 2024. According to whether the nodule margin was ≤ 2 mm from the thyroid capsule，the patients were divided into the subcapsular group（60 cases）and the remote from the capsule group（70 cases）. The clinical data，ablation parameters，complication rate，absorption of ablation lesions，tumor recurrence and metastasis were compared between the two groups.Results:The success rates of RFA in both groups were 100%. There was a statistically significant difference between the two groups at baseline in the location of nodules（ χ2=13.266， P=0.001）. After 12 months of routine follow-up，the incidence of complications was 5.00%（3/60）in the subcapsular group and 1.43%（1/70）in the remote from the capsule group，all complications recovered within 1 month，with no statistically significant difference（ P=0.407）. There were no significant changes in thyroid function［including free triiodothyronine（FT3），free thyroxine（FT4）and thyroid-stimulating hormone（TSH）］before and one month after RFA in the two groups（all P>0.05）. A statistically significant difference in the volume reduction rate（VRR）between the two groups was observed only when maximum diameter ≤ 5?mm at the 1st and 3rd months（all P<0.05）. For the remaining postoperative follow-up periods，there were no statistically significant differences in ablation zone volume or VRR between the two groups（all P>0.05）. At the final follow-up，for lesions with maximum diameter ≤ 5?mm，the tumor disappearance rate was 100% in both the subcapsular and remote from the capsule groups. For lesions with maximum diameter >5 mm，the rates were 78.4%（29/37）and 89.7%（26/29）in the respective groups，with no statistically significant difference between the two groups（ χ2 = 1.489， P =0.222）.There was no local tumor progression，lymph node metastasis or distant metastasis in either group. Conclusions:RFA is a safe and effective treatment for subcapsular PTMC，demonstrating comparable efficacy to that for PTMC located remote from the thyroid capsule.

Objective:To evaluate the safety and efficacy of radiofrequency ablation（RFA）in the treatment of papillary thyroid microcarcinoma（PTMC）near and remote from the thyroid capsule.Methods:The clinical data，ultrasound images，ablation parameters and postoperative follow-up metrics were prospectively collected from 130 patients with pathologically confirmed PTMC in the Second Affiliated Hospital of Xi'an Jiaotong University who underwent RFA treatment between January 2023 and December 2024. According to whether the nodule margin was ≤ 2 mm from the thyroid capsule，the patients were divided into the subcapsular group（60 cases）and the remote from the capsule group（70 cases）. The clinical data，ablation parameters，complication rate，absorption of ablation lesions，tumor recurrence and metastasis were compared between the two groups.Results:The success rates of RFA in both groups were 100%. There was a statistically significant difference between the two groups at baseline in the location of nodules（ χ2=13.266， P=0.001）. After 12 months of routine follow-up，the incidence of complications was 5.00%（3/60）in the subcapsular group and 1.43%（1/70）in the remote from the capsule group，all complications recovered within 1 month，with no statistically significant difference（ P=0.407）. There were no significant changes in thyroid function［including free triiodothyronine（FT3），free thyroxine（FT4）and thyroid-stimulating hormone（TSH）］before and one month after RFA in the two groups（all P>0.05）. A statistically significant difference in the volume reduction rate（VRR）between the two groups was observed only when maximum diameter ≤ 5?mm at the 1st and 3rd months（all P<0.05）. For the remaining postoperative follow-up periods，there were no statistically significant differences in ablation zone volume or VRR between the two groups（all P>0.05）. At the final follow-up，for lesions with maximum diameter ≤ 5?mm，the tumor disappearance rate was 100% in both the subcapsular and remote from the capsule groups. For lesions with maximum diameter >5 mm，the rates were 78.4%（29/37）and 89.7%（26/29）in the respective groups，with no statistically significant difference between the two groups（ χ2 = 1.489， P =0.222）.There was no local tumor progression，lymph node metastasis or distant metastasis in either group. Conclusions:RFA is a safe and effective treatment for subcapsular PTMC，demonstrating comparable efficacy to that for PTMC located remote from the thyroid capsule.

Objective:To explore the diagnostic value of different machine learning models and optimal machine learning model combined with clinical data for diagnosing Hashimoto′s thyroiditis (HT).Methods:The thyroid gland images of 643 patients with 643 thyroid nodules who underwent preoperative ultrasound examination and had pathological results in the Second Affiliated Hospital of Xi′an Jiaotong University from December 2018 to March 2024 were retrospectively collected, and the images were divided into training set and test set according to a ratio of 7 to 3. Twenty ultrasound imaging omics models were constructed using pairwise combination of 5 feature screening components and 4 classifiers. The area under the curve (AUC) of each model in the test set was compared. Meanwhile, 3 basic network models were respectively used to construct deep learning models for diagnosing HT, and the diagnostic efficacies of the deep learning models and the ultrasound imaging omics models for HT were compared. The model with the greatest efficacy was selected as the optimal machine learning model. Further, the optimal machine learning model was combined with clinical data to construct a combined model. The ROC curves were plotted to compare the diagnostic efficacy of the optimal machine learning model and the combined model for HT.Results:In the comparison of the efficacies of ultrasound imaging omics models and deep learning models in diagnosing HT, the efficacy of stable feature screening-logistic regression (LR) model was the greatest, and the accuracy, sensitivity and specificity of using the LR model in diagnosing HT in the test set were 78%, 75%, 74%, respectively, with an AUC of 0.82(95% CI=0.76-0.88). After combining the LR model with clinical data, the accuracy, sensitivity, and specificity of the combined model in the test set were 87%, 74%, and 95%, respectively, with an AUC of 0.91(95% CI=0.87-0.95), which was strongly consistent with pathology (Kappa value=0.708, P<0.001). Conclusions:The optimal machine learning model (LR model) constructed in this study demonstrates a strong ability to diagnose HT and can accurately detect patients with atypical ultrasound manifestations of HT. The combination with clinical data can improve its diagnostic efficacy with higher accuracy and specificity.

Objective:To explore the diagnostic value of different machine learning models and optimal machine learning model combined with clinical data for diagnosing Hashimoto′s thyroiditis (HT).Methods:The thyroid gland images of 643 patients with 643 thyroid nodules who underwent preoperative ultrasound examination and had pathological results in the Second Affiliated Hospital of Xi′an Jiaotong University from December 2018 to March 2024 were retrospectively collected, and the images were divided into training set and test set according to a ratio of 7 to 3. Twenty ultrasound imaging omics models were constructed using pairwise combination of 5 feature screening components and 4 classifiers. The area under the curve (AUC) of each model in the test set was compared. Meanwhile, 3 basic network models were respectively used to construct deep learning models for diagnosing HT, and the diagnostic efficacies of the deep learning models and the ultrasound imaging omics models for HT were compared. The model with the greatest efficacy was selected as the optimal machine learning model. Further, the optimal machine learning model was combined with clinical data to construct a combined model. The ROC curves were plotted to compare the diagnostic efficacy of the optimal machine learning model and the combined model for HT.Results:In the comparison of the efficacies of ultrasound imaging omics models and deep learning models in diagnosing HT, the efficacy of stable feature screening-logistic regression (LR) model was the greatest, and the accuracy, sensitivity and specificity of using the LR model in diagnosing HT in the test set were 78%, 75%, 74%, respectively, with an AUC of 0.82(95% CI=0.76-0.88). After combining the LR model with clinical data, the accuracy, sensitivity, and specificity of the combined model in the test set were 87%, 74%, and 95%, respectively, with an AUC of 0.91(95% CI=0.87-0.95), which was strongly consistent with pathology (Kappa value=0.708, P<0.001). Conclusions:The optimal machine learning model (LR model) constructed in this study demonstrates a strong ability to diagnose HT and can accurately detect patients with atypical ultrasound manifestations of HT. The combination with clinical data can improve its diagnostic efficacy with higher accuracy and specificity.

Objective:To compare and analyze the clinical diagnostic values of five thyroid nodule malignant risk stratification guidelines.Methods:From October 2019 to October 2021, 926 cases of patients with 1 027 thyroid nodules were recruited in the Second Affiliated Hospital of Xi ′an Jiaotong University. All nodules were categorized individually according to 2015 American Thyroid Association for Ultrasound Malignancy Risk Stratification of Thyroid Nodules in Adults Guidelines(ATA guidelines), 2016 the Korean Society of Radiology and the Korean Society of Thyroid Radiology Thyroid Imaging Reporting and Data Systems(K-TIRADS), 2017 European Thyroid Association Thyroid Imaging Reporting and Data Systems(Eu-TIRADS), 2017 American College of Radiology Thyroid Imaging Reporting and Data Systems (ACR TI-RADS), and 2020 Chinese Thyroid Imaging Reporting and Data System (C-TIRADS). The pathological results were taken as the gold standard, the malignancy ratio of nodules of different categories in each system was calculated. ROC curves were plotted to evaluate the diagnostic efficiencies of different systems, and DeLong test was used to compare the areas under ROC curves. The sensitivity and specificity of different systems were calculated when the maximum point of the Youden index was the optimal cut-off value.Results:In the same stratified system, there were statistically significant differences in the malignant proportion of nodules of different grades ( P<0.05). The malignant proportion of nodules in the high-risk group showed no significant difference among different stratified systems ( P>0.05). Except for C-TIRADS, the malignant proportion of nodules was increased with the increase of diagnostic grade at each grade of the other four stratification systems. ROC curve showed that AUCs of ATA guidelines, K-TIRADS, EU-TIRADS, ACR TI-RADS and C-TIRADS were 0.814, 0.819, 0.814, 0.820 and 0.802, respectively, there was no statistical significance in AUC of different stratification systems (all P>0.05). The optimal truncation values in differentiating benign and malignant nodules were middle-risk malignant nodules, moderately suspicious malignant nodules, middle-risk malignant nodules, class 4 and class 4B. The diagnostic of five stratification systems showed that ATA guidelines had the highest sensitivity (0.784), C-TIRADS had the highest specificity (0.854). Conclusions:The five stratified systems have similar efficacy in differentiating benign and malignant thyroid nodules, and all of them have good diagnostic value.