Clinical application value of nomogram model based on clinical and ultrasound features in predicting thyroid C-TI-RADS classification optimization

doi:10.21203/rs.3.rs-2010247/v1

Backgrounds: A nomogram model based on clinical and ultrasound features was constructed to explore its clinical application value in predicting thyroid C-TI-RADS classification optimization.

Methods: Clinical data and ultrasound imaging data of 1,234 patients with thyroid nodules collected from January 2021 to February 2022 of Sichuan Provincial People's Hospital were retrospectively analyzed.All patients underwent preoperative thyroid ultrasound examination and retained standard ultrasound images, evaluated the thyroid nodule C-TI-RADS classification, using the postoperative pathological results as the "gold standard". Independent predictors of C-TI-RADS classification optimization were selected by univariate and multivariate logstic regression analysis, and a nomogram prediction model(*C-TI-RADS) was constructed.The internal validation of the model was performed by Bootstrap resampling. ROC curve was drawn to evaluate the discrimination of the model, and calibration curve and decision curve were drawn to evaluate the consistency and clinical practicability of the prediction model.

Results: C-TI-RADS classification, size and number of thyroid nodules, abnormal cervical lymph node ultrasonography, sex and age were independent factors for predicting C-TI-RADS classification optimization (all P < 0.05).The C index of the nomogram prediction model(*C-TI-RADS) constructed based on the above factors was 0.790 (95%CI: 0.765–0.815).Under the optimal cut-off value, the sensitivity was 70.8%, the specificity was 74.4%, and the accuracy was 72.2%.The calibration curve and decision curve showed good consistency and clinical practicability of the model.

Conclusions: Nomogram model has good accuracy in the prediction of thyroid C-TI-RADS classification optimization, and can assist ultrasound physician to modify C-TI-RADS classification, which has potential clinical application value.

Thyroid

TI-RADS

Optimization

Nomogram

Prediction model

Thyroid Imaging Reporting and Data System (TI-RADS) is widely used in clinical practice. The classification was assessed based on the imaging characteristics of thyroid nodules^[1–3], and it does not involve clinical risk factors of patients and subjective judgment of sonographers. While, in 2021, the Chinese Thyroid Imaging Reporting and Data System (C-TI-RADS) guidelines clearly proposed that: The sonographer may modify the C-TI-RADS classification level (up or down) based on the physician's experience, medical history, and specific types of ultrasound images. Each risk level of TI-RADS has corresponding a clinical management mean.

According to C-TI-RADS guidelines^[3], for thyroid nodules 4B or above, whether to perform FNA should consider the size of the lesion, whether there are multiple lesions and the location of the lesion, etc. For 4B single-focal nodules smaller than 10mm, which are not adjacent to the trachea, capsule or recurrent laryngeal nerve, active monitoring can be selected. For 4A nodules, FNA is still recommended if the nodule is larger than 15mm; For nodules less than or equal to level 3, FNA is not required. In addition, the C-TI-RADS guidelines indicate that ultrasonographic evaluation of cervical lymph nodes can provide important information for risk stratification of thyroid nodules. When the cervical lymph nodes showed metastatic signs and there are suspicious nodules in the thyroid gland, the nodules could be evaluated into level 5. The ultrasonographic features of cervical lymph node metastases include roundness, microcalcification, cystic changes, focal hyperecho, and local or diffuse hyperblood supply. In addition, in young patients, papillary thyroid carcinoma tends to be active and often actively treated in clinical practice, while in elderly patients, papillary thyroid carcinoma tends to develop slowly, which is suitable for conservative treatment and active monitoring^[4]. Therefore, sonographers can appropriately modify TI-RADS classification according to additional ultrasound information of thyroid nodules and clinical data of patients. However, there is no unified standard on how to modify TI-RADS classification ^{[3, 5]}. It is an urgent problem to establish an objective and scientific TI-RADS optimization prediction model in line with clinical practice, so as to improve the consistency of TI-RADS classification and assist clinicians in formulating individualized treatment plans. This study intends to establish a nomogram model for predicting thyroid C-TI-RADS optimization based on the classification level of C-TI-RADS, size and number of thyroid nodules, ultrasound manifestations of cervical lymph nodes, sex and age of patients, and explore its clinical application value.

Patients

A total of 1234 patients who underwent thyroid nodule resection in Sichuan Provincial People's Hospital from January 2021 to February 2022 were enrolled. Inclusion criteria: ① The pathological results of thyroid nodules were clear; ② Preoperative ultrasound images of thyroid nodules and cervical lymph nodes were clear; ③ Ultrasound records were complete (including the size and number of thyroid nodules, and whether cervical lymph nodes were abnormal); ④Accurate clinical information of patients. Exclusion criteria: ① The pathological results of thyroid nodules were not clear; ② Thyroid biopsy without surgical resection; ③ Poor preoperative ultrasound image.

Operation Approach

The clinical data and thyroid ultrasound imaging data of the patients were collected and recorded by 2 primary ultrasound physicians using a structured data sheet customized for this study. The clinical data included sex and age of the patients. Ultrasound imaging data include thyroid nodule composition, echo, calcification, margin, aspect ratio, thyroid nodule size, number, cervical lymph nodes, etc^[6]. The two ultrasound primary physicians independently performed C-TI-RADS classification by counting method ^[5] (see Table 1). The counting criteria included five ultrasound features, such as solid, very hypoechoic, punctate strong echo, blurred edge, and vertical position. Subsequently, the C-TI-RADS classification was independently optimized by two intermediate ultrasound physicians considering the sex and age of the patients, the size and number of nodules, and the ultrasound characteristics of cervical lymph nodes. The optimized C-TI-RADS classification was named *C-TI-RADS. All images were analyzed without knowing the pathological results of thyroid nodules, and consensus results were obtained through negotiation when there were differences. When the level of C-TI-RADS classification is inconsistent with that of *C-TI-RADS classification, the level of C-TI-RADS classification should be optimized, including classification improvement or classification reduction .

Statistical analysis

SPSS 26.0 and R Studio software were used for statistical analysis. Enumeration data were expressed as the number of cases (N) or percentage (%). χ² test or Fisher exact test were used for comparison between groups. Univariate logstic regression analysis was used to screen the risk factors with statistically significant differences, and then multivariate logistic regression was used to analyze the correlation between these risk factors and the prediction of C-TI-RADS optimization, and the prediction model was established. ROC curve was drawn to evaluate the discrimination of the prediction model, and Bootstrap method was used to resampling 1000 times for internal validation of the prediction model. The calibration curve was drawn to evaluate the calibration degree of the model, and the clinical decision curve was drawn to evaluate the clinical practicability of the model. P < 0.05 (two-sided test) was considered statistically significant.

The baseline information

A total of 1234 patients were enrolled in this study, with an average age of (44.34 ± 13.08) years, including 354 males (28.7%) and 880 females (71.3%). The incidence of C-TI-RADS classification optimization was 61.7% (761/1234), and level 4C (36.5%) accounted for the highest proportion of all the optimization, followed by level 4B (12.5%). The average size of thyroid nodules was (15.73 ± 11.80) mm. There were 575 cases (46.5%) of small nodules ≤ 10mm, 703 cases (57.0%) of single focal nodules, and 273 cases (22.1%) of abnormal cervical lymph node ultrasound. There were significant differences between C-TIRAS and *C-TI-RADS (P < 0.05), and the diagnostic efficacy of the two classifications for benign and malignant thyroid nodules was different (P < 0.05). See Table 2 and Table 3.

Univariate And Multivariate Logistic Regression Analysis

Univariate logstic regression analysis showed that there were statistically significant differences in sex and age, thyroid nodule size, cervical lymph node ultrasound abnormalities and C-TI-RADS classification level. Although there was no significant difference between single and multifocal thyroid nodules (P > 0.05), the number of nodules is still a risk factor for C-TI-RADS classification in clinical practice. Therefore, the above six predictors were included in the multivariate logstic regression analysis, and the results showed that all of them were independent predictors of C-TI-RADS classification optimization (P < 0.05). For example, the probability of optimizing the C-TI-RADS classification is increased when any of the following conditions occur, the C-TI-RADS classification is 4B, the patient is male, age ≤ 60 years old, thyroid nodule is multifocal, nodule is ≤ 30mm, or cervical lymph node ultrasound is abnormal. See Table 4.

Construction And Evaluation Of Nomogram Model

A nomogram prediction model for C-TI-RADS classification optimization was constructed based on the six independent predictors, including C-TI-RADS classification, patient sex, age, thyroid nodule size, number, and cervical lymph node ultrasound abnormality, the nomogram prediction model was the * C-TI-RADS,as shown in Fig. 1a. In the nomogram model, each predictor corresponds to a score, and when each score is added, the total score corresponds to a percentage, which represents the probability of C-TI-RADS classification optimization. The C index of the prediction model was 0.790 (95%CI: 0.765–0.815). The area under the ROC curve (AUC) was 0.79 (Fig. 2d), and the optimal cut-off value for predicting C-TI-RADS classification optimization was 0.637, the corresponding sensitivity was 70.8%, the specificity was 74.4%, and the accuracy was 72.2%. Hosmer⁃Lemeshow test showed P = 0.37, indicating that the model had a good degree of fit. Bootstrap method was used to re-sample 1000 times for internal verification, and showed that the nomogram model and its correction model were close to the ideal curve (Fig. 2e), indicating that the calibration degree of the model was good. The clinical decision curve suggested that the net clinical benefit obtained by using nomogram model to selectively optimize C-TI-RADS was higher than that obtained by optimizing all thyroid nodules or not optimizing all thyroid nodules (Fig. 2f), indicating that the model had high clinical practicability. In the diagnostic efficiency curve, the AUC of *C-TI-RADS (0.969) was greater than that of C-TI-RADS (0.947) (Fig. 2g), indicating that the diagnostic efficiency of thyroid nodules was improved after optimization.

Example Of Nomogram Application

A 65-year-old male patient presented with a single focal thyroid nodule of 7.8mm. The ultrasound image features of thyroid nodules were solid, hypoechoic, aspect ratio greater than 1, and edge underclarity. According to C-TI-RADS classification (see Table 1), the count of risk factors was 4, so it was judged as 4C level (Fig. 1b). In addition, there were abnormal cervical lymph nodes (Fig. 1c). Finally, according to the nomogram prediction model, the overall score of this thyroid nodule was 157, corresponding to an optimization probability of greater than 80% (Fig. 1a). After optimization, the ultrasound report suggested TI-RADS level 5.

C-TI-RADS classification is based on the ultrasound image characteristics of thyroid nodules, and does not involve the clinical data of patients or other ultrasound information. However, factors such as sex and age of patients, size and number of thyroid nodules and cervical lymph node abnormality or not affect the clinical treatment of thyroid nodules ^[7–9]. At present, the optimization of thyroid nodule classification is based on the subjective experience of sonogrphers, and there is no unified optimization standard. Therefore, we developed and internally validated a clinical prediction model for predicting C-TI-RADS classification optimization.

In this study, according to Logisitic regression analysis, C-TI-RADS classification level, cervical lymph node ultrasound abnormality and thyroid nodule size were the most important predictors of C-TI-RADS classification optimization, followed by sex, age and number of thyroid nodule. The established nomogram model for predicting C-TI-RADS classification optimization has an AUC of 0.790, a sensitivity of 70.8%, a specificity of 74.4% and an accuracy of 72.2%. The calibration curve and clinical decision curve also showed good consistency and net benefit, indicating that the prediction model has potential clinical application value.

A study of 5162 healthy individuals followed up for 5 years ^[10] showed that most nodules of level 4A and above had their classification adjustment during the follow-up, and the higher the classification level, the higher the risk of malignant progression. The results of this study showed that in C-TI-RADS classification, 4A/4B/4C had a higher probability of optimization (72.1%, 75.1% and 69.0%, respectively), and the OR value of 4B was 33.1, which was consistent with the clinical consensus that 4B was used as the cut-off point for benign and malignant diagnosis of thyroid nodules^[3]. Besides, 85.3% of the patients with abnormal cervical lymph node ultrasonography underwent C-TI-RADS classification optimization. In multivariate regression analysis, the adjusted OR value of cervical lymph node ultrasound abnormalities was 5.0, which further confirmed that cervical lymph node ultrasound abnormality was an independent risk factor for effectively predicting the occurrence of C-TI-RADS classification optimization, and could be used as a supplementary diagnostic factor for thyroid nodules. In terms of thyroid nodule size, the prediction model showed that thyroid micronodules (less than 10mm) were more likely to have C-TI-RADS classification optimization than thyroid macronodules (the percentage of optimization was 61.2% and 50.9%, respectively). For non-minor nodules with high C-TI-RADS level (such as 4B), needle biopsy or surgery is the best treatment. However, most of the thyroid micronodules with higher classification level almost are small carcinomas with low malignant degree ^[11] and they often growth slowly. Sonographers tend to reduce their classification in actual diagnosis, so as to reduce unnecessary needle biopsy or surgery of thyroid micronodules and avoid overtreatment. In addition, in terms of patient age, the prediction model showed that young men had a higher probability of C-TI-RADS classification optimization than older women. Many studies have also proposed that young male patients with thyroid cancer are more malignant, aggressive and have a greater risk of recurrence ^{[12, 13]}.

Compared with other TI-RADS classification optimization models, this study included clinical risk factors that considered in the actual diagnosis, which was in line with the clinical diagnosis model of thyroid nodule. At present, most prediction studies ^[14–16] establish prediction models based on ultrasonic characteristics of thyroid nodules, including location, size, composition, echo, shape, margin, calcification, blood flow signal, halo sign, etc. Although the ultrasound features included in various studies are different, overall, the optimized TI-RADS classification can maintain high sensitivity and diagnostic accuracy, and avoid some unnecessary biopsies or surgeries ^[17]. Ling Chen .et al ^[18] included nodular halo sign and age as risk factors in their prediction model, and obtained high sensitivity (88.4%) and specificity (91.7%).In addition, some predictive models also take such as ultrasound elastography, enhanced ultrasound, genetic testing and laboratory indicators into account^[19–22], which are more accurate and sensitive than single TI-RADS classification. However, whether the improvement of its diagnostic value is greater than the improvement of cost of the medical treatment still needs in-depth clinical research to verify. While, the prediction model in this study not only improves the diagnostic efficiency of benign and malignant thyroid nodules (AUC = 0.969), but also does not involve other imaging or biochemical examination techniques, which is in line with the principles of health economics and has direct clinical application value.

This study has the following limitations: ① Inter-observer variability and the subjectivity of recognition images may affect the validity of the model. ② This study is a single center retrospective study, and there may be some selection bias. ③ This study discussed the risk of C-TI-RADS classification optimization, but did not discuss whether the result of optimization increased or decreased. ④ other potential clinical risk factors were not included in this study, such as body mass index, living environment, medical environment, location of thyroid nodules, and radiation exposure history of patients ^[23–25]. In the future, large sample studies are still needed to further verify the validity and extrapolation of this model, and to explore multi-center and prospective prediction models with higher diagnostic efficiency.

In conclusion, the nomogram prediction model in this study integrates the clinical data of patients and the ultrasonic signs of thyroid nodules, showing high accuracy and discrimination in the prediction of C-TI-RADS classification optimization, which can assist ultrasound physician to achieve consistency in modified C-TI-RADS classification, and has potential clinical application value.

C-TI-RADS: Chinese Thyroid Imaging Reporting and Data System; FNA: Fine needle aspiration; ROC: Receiver Operating Characteristic; AUC: Area under curve.

Acknowledgements

We thank LinXian Yue for his help with the revision of manuscript.

Authors’ contributions

YL ,TX and QC designed the study and drafted the manuscript. JS, XL and EXF collected the data. JZ and FHH performed statistical analysis. All authors read and approved the final manuscript.

Funding

This study was not supported by a fund.

Availability of data and materials

All data analyzed during this study are available from the corresponding author on reasonable request.

Ethics approval and consent to participate

This study protocol was approved by the Ethics Committee of Sichuan Academy of Medical Sciences and Sichuan Provincial People's Hospital (Approval number: Ethics Approval (Research) No. 277, 2022).All data were completely anonymous, and the requirement for informed consent was waived by our ethics committee.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Haugen BR, Alexander EK, Bible KC, Doherty GM, Mandel SJ, Nikiforov YE, Pacini F, Randolph GW, Sawka AM, Schlumberger M, Schuff KG, Sherman SI, Sosa JA, Steward DL, Tuttle RM, Wartofsky L. 2015 American Thyroid Association Management Guidelines for Adult Patients with Thyroid Nodules and Differentiated Thyroid Cancer: The American Thyroid Association Guidelines Task Force on Thyroid Nodules and Differentiated Thyroid Cancer. Thyroid. 2016 ;26(1):1-133.
Tessler FN, Middleton WD, Grant EG, Hoang JK, Berland LL, Teefey SA, Cronan JJ, Beland MD, Desser TS, Frates MC, Hammers LW, Hamper UM, Langer JE, Reading CC, Scoutt LM, Stavros AT. ACR Thyroid Imaging, Reporting and Data System (TI-RADS): White Paper of the ACR TI-RADS Committee. J Am Coll Radiol. 2017 ;14(5):587-95.
Superficial Organ and Vascular Ultrasound Group, Society of Ultrasound in Medicine, Chinese Medical Association; Chinese Artificial Intelligence Alliance for Thyroid and Breast Ultrasound.2020 Chinese Guidelines for Ultrasound Malignancy Risk Stratification of Thyroid Nodules: The C-TIRADS.Chin J Ultrasonogr. 2021;30(3): 185-200.
Zhou J, Yin L, Wei X, Zhang S, Song Y, Luo B, Li J, Qian L, Cui L, Chen W, Wen C, Peng Y, Chen Q, Lu M, Chen M, Wu R, Zhou W, Xue E, Li Y, Yang L, Mi C, Zhang R, Wu G, Du G, Huang D, Zhan W; Superficial Organ and Vascular Ultrasound Group of the Society of Ultrasound in Medicine of the Chinese Medical Association; Chinese Artificial Intelligence Alliance for Thyroid and Breast Ultrasound. 2020 Chinese guidelines for ultrasound malignancy risk stratification of thyroid nodules: the C-TIRADS. Endocrine. 2020 ;70(2):256-79.
Sugitani I, Ito Y, Takeuchi D, Nakayama H, Masaki C, Shindo H, Teshima M, Horiguchi K, Yoshida Y, Kanai T, Hirokawa M, Hames KY, Tabei I, Miyauchi A. Indications and Strategy for Active Surveillance of Adult Low-Risk Papillary Thyroid Microcarcinoma: Consensus Statements from the Japan Association of Endocrine Surgery Task Force on Management for Papillary Thyroid Microcarcinoma. Thyroid. 2021;31(2):183-92.
Expert Consensus on some common clinical problems of thyroid and related cervical lymph node ultrasound (2018 edition). Chin J Ultrasound Med.2019;35(03):193-204.
Apostolou K, Zivaljevic V, Tausanovic K, Zoric G, Chelidonis G, Slijepcevic N, Jovanovic M, Paunovic I. Prevalence and risk factors for thyroid cancer in patients with multinodular goitre. BJS Open. 2021;5(2):zraa014.
Kauffmann RM, Hamner JB, Ituarte PHG, Yim JH. Age greater than 60 years portends a worse prognosis in patients with papillary thyroid cancer: should there be three age categories for staging? BMC Cancer. 2018;18(1):316.
Joseph KR, Edirimanne S, Eslick GD. Multifocality as a prognostic factor in thyroid cancer: A meta-analysis. Int J Surg. 2018;50:121-25.
Zahedi A, Bondaz L, Rajaraman M, Leslie WD, Jefford C, Young JE, Pathak KA, Bureau Y, Rachinsky I, Badreddine M, De Brabandere S, Fong H, Maniakas A, Van Uum S. Risk for Thyroid Cancer Recurrence Is Higher in Men Than in Women Independent of Disease Stage at Presentation. Thyroid. 2020;30(6):871-77.
Grønlund MP, Jensen JS, Hahn CH, Grønhøj C, Buchwald CV. Risk Factors for Recurrence of Follicular Thyroid Cancer: A Systematic Review. Thyroid. 2021;31(10):1523-30.
Xu S, Han Y. The Overdiagnosis of Thyroid Micropapillary Carcinoma: The Rising Incidence, Inert Biological Behavior, and Countermeasures. J Oncol. 2021 Jul 9;2021:5544232.
Liang Yu, Yue Linxian, Chen Qin, Lin Jie, Guo Daoning, He Peng. Efficiency comparison of Kwak and ACR (2017) Thyroid Imaging Reporting and Data System(TI-RADS) classification: a polycentric retrospective study. Chin J Ultrasonogr. 2019;28(5): 419-24.
Xin Y, Liu F, Shi Y, Yan X, Liu L, Zhu J. A Scoring System for Assessing the Risk of Malignant Partially Cystic Thyroid Nodules Based on Ultrasound Features. Front Oncol. 2021;11:731779.
Xiao J, Xiao Q, Cong W, Li T, Ding S, Shao C, Zhang Y, Liu J, Wu M, Jia H. Discriminating Malignancy in Thyroid Nodules: The Nomogram Versus the Kwak and ACR TI-RADS. Otolaryngol Head Neck Surg. 2020 ;163(6):1156-65.
Zhu JL, Zhao J, Wei X, Zhang S. Value of modified TI-RADS in the diagnosis of medullary thyroid carcinoma. Zhonghua Zhong Liu Za Zhi. 2017;39(8):618-23.
Na DG, Paik W, Cha J, Gwon HY, Kim SY, Yoo RE. Diagnostic performance of the modified Korean Thyroid Imaging Reporting and Data System for thyroid malignancy according to nodule size: a comparison with five society guidelines. Ultrasonography. 2021;40(4):474-85.
Jin ZQ, Yu HZ, Mo CJ, Su RQ. Clinical Study of the Prediction of Malignancy in Thyroid Nodules: Modified Score versus 2017 American College of Radiology's Thyroid Imaging Reporting and Data System Ultrasound Lexicon. Ultrasound Med Biol. 2019;45(7):1627-37.
Huang X, Wu Z, Zhou A, Min X, Qi Q, Zhang C, Chen S, Xu P. Nomogram Combining Radiomics With the American College of Radiology Thyroid Imaging Reporting and Data System Can Improve Predictive Performance for Malignant Thyroid Nodules. Front Oncol. 2021;11:737847.
Guo SY, Zhou P, Zhang Y, Jiang LQ, Zhao YF. Exploring the Value of Radiomics Features Based on B-Mode and Contrast-Enhanced Ultrasound in Discriminating the Nature of Thyroid Nodules. Front Oncol. 2021;11:738909.
Shi Hongwei,Wang Huiyang, Deng Huadong,Peng Chenbing, Jiang Tianan. Diagnostic value of contrast-enhanced ultrasound combined with fine-needle aspiration biopsy and BRAF gene detection for TI-RADS category 4 nodules.Chin J Ultrasonogr. 2022; 31(3): 214-19.
Li M, Wei L, Li F, Kan Y, Liang X, Zhang H, Liu J. High Risk Thyroid Nodule Discrimination and Management by Modified TI-RADS. Cancer Manag Res. 2021;13:225-234.
Chen L, Zhang J, Meng L, Lai Y, Huang W. A new ultrasound nomogram for differentiating benign and malignant thyroid nodules. Clin Endocrinol (Oxf). 2019;90(2):351-59.
Goundan PN, Mamou J, Rohrbach D, Smith J, Patel H, Wallace KD, Feleppa EJ, Lee SL. A Preliminary Study of Quantitative Ultrasound for Cancer-Risk Assessment of Thyroid Nodules. Front Endocrinol (Lausanne). 2021;12:627698.
Zhang F, Zheng B, Yu X, Wang X, Wang S, Teng W. Risk Factors for Contralateral Occult Carcinoma in Patients With Unilateral Papillary Thyroid Carcinoma: A Retrospective Study and Meta-Analysis. Front Endocrinol (Lausanne). 2021;12:675643.

Table 1. Chinese Thyroid Imaging Reporting and Data system

Classification	Value	Evaluation	Risk of malignancy （%）	Ultrasonographic findings and value criteria
1	NA	No nodule	0	Positive indicators： Aspect ratio > 1（+1） Solid（+1） Very low echo（+1） Punctate strong echo（Suspicious microcalcification）（+1） Blurred or irregular edges/ Extra-thyroidal extension（+1） Negative indicators： Punctate strong echo（Comet Tail artifact）（-1）
2	-1	Benign	0
3	0	May be benign	<2
4A	1	Low suspected malignancy	2~10
4B	2	Medium suspected malignancy	10~50
4C	3~4	Highly suspected maglinancy	50~90
5	5	Highly suggestive maglinancy	＞90
6		Malignancy was confirmed by biopsy

Table 2. Baseline characteristics of enrolled patients [Example (%）]

Variate		C-TI-RADS classification optimization			P
		Total（n=1234）	Optimization（n=761）	No optimization（n=473）
Sex					0.016
	Female	880(71.3)	524(42.5)	356(28.8)
	Male	354(28.7)	237(19.2)	117(9.5)
Age(year)					＜0.001
	≤40	524（42.5）	343（27.8）	181（14.7）
	40~60	563（45.6）	348（28.2）	215（17.4）
	＞60	147（11.9）	70（5.7）	77（6.2）
Nodular size（mm）					0.003
	≤10	575(46.6)	352(28.5)	223(18.1)
	10~30	496(40.2)	326(26.4)	170(13.8)
	＞30	163(13.2)	83(6.7)	80(6.5)
Nodular number					0.513
	Single	703(57.0)	428(34.7)	275(22.3)
	Multiple	531(43.0)	333(27.0)	198(16.0)
Substance					＜0.001
	Mix	153(12.4)	32(2.6)	121(9.8)
	Solid	1065(86.3)	727(58.9)	338(27.4)
	Cystic	16（1.3）	2(0.2)	14(1.1)
Echo					0.878
	Equal or High	116（9.4）	69（5.6）	47（3.8）
	Low	763（61.8）	472（38.2）	291（23.6）
	Very low	355（28.8）	220（17.8）	135（10.9）
Calcification					0.002
	None	551(44.7)	310(25.1)	241(19.5)
	Microcalciffcation	620(50.2)	409(33.1)	211(17.1)
	Coarse calcification	63(5.1)	42(3.4)	21(1.7)
Edge blur					＜0.001
	No	477(38.7)	257(20.8)	220(17.8)
	Yes	757(61.3)	504(40.8)	253(20.5)
Aspect ratio > 1					0.075
	No	876(71.0)	554(44.9)	322(26.1)
	Yes	358(29.0)	207(16.8)	151(12.2)
Abnormal of cervical lymph nodes					＜0.001
	No	961(77.9)	528(42.8)	433(35.1)
	Yes	273(22.1)	233(18.9)	40(3.2)
C-TI-RADS classification					＜0.001
	3 category	150(12.2)	23(1.9)	127(10.3)
	4A category	151(12.2)	109(8.8)	42(3.4)
	4B category	205(16.6)	154(12.5)	51(4.1)
	4C category	652(52.8)	450(36.5)	202(16.4)
	5 category	76(6.2)	25(2.1)	51(4.1)

Table 3. C-TI-RADS / * C-TI-RADS of Thyroid nodule and Pathological results [Example]

	C-TI-RADS（before optimization）		*C-TI-RADS（after optimization）
	Benign	Malignancy	Benign	Malignancy
3 category	145	5	188	6
4A category	90	61	64	49
4B category	31	174	18	186
4C category	10	642	4	313
5 category	0	76	2	404
	χ2=788.534,P＜0.001		χ2=911.961,P＜0.001

Table 4. Analysis of clinical risk factors for predicting C-TI-RADS optimization

Variate	Univariate logistic analysis		Multivariate logistic analysis
	*OR（95%CI）*	P	*OR（95%CI）*	P
Sex（Male vs Female）	1.376(1.062~1.783)	0.016	1.526(1.129~2.062)	0.006
Age (year)
≤40	2.085(1.439~3.019)	0	1.419(0.891~2.260)	0.141
40~60	1.780(1.235~2.566)	0.002	1.751(1.115~2.749)	0.015
＞60	1		1
Nodular size（mm）
≤10	1.521(1.072~2.159)	0.019	0.381(0.207~0.700)	0.002
10~30	1.848(1.291~2.646)	0.001	0.483(0.268~0.871)	0.016
＞30	1		1
Nodular number（Multiple vs Single）	1.081(0.857~1.363)	0.513	1.809(1.359~2.409)	0
Substance
Mix	1.851(0.400~8.566)	0.431
Solid	15.056(3.403~66.619)	0
Cystic	1
Echo
Equal or High	1
Low	1.105(0.742~1.646)	0.624
Very low	1.11(0.723~1.703)	0.633
Calcification
None
Microcalcification	1.507(1.190~1.909)	0.001
Coarse calcification	1.555(0.897~2.695)	0.116
Edge blur（Yes vs No）	0.586(0.464~0.742)	0
Aspect ratio > 1（Yes vs No）	0.797(0.620~1.024)	0.076
Abnormal cervical lymph nodes （Yes vs No）	4.777(3.337~6.838)	0	5.004(3.302~7.583)	0
C-TI-RADS classification
3 category	1		1
4A category	14.330(8.111~25.319)	0	26.851(13.749~52.437)	0
4B category	16.673(9.663~28.769)	0	33.129(16.554~66.300)	0
4C category	12.301(7.656~19.763)	0	20.055(10.492~38.337)	0
5 category	2.707(1.409~5.200)	0.003	3.575(1.569~8.145)	0.002

renamed2df6f.xlsx

Clinical application value of nomogram model based on clinical and ultrasound features in predicting thyroid C-TI-RADS classification optimization

Status:

Version 1

Abstract

Figures

Background

Methods

Patients

Operation Approach

Statistical analysis

Results

The baseline information

Univariate And Multivariate Logistic Regression Analysis

Construction And Evaluation Of Nomogram Model

Example Of Nomogram Application

Discussion

Conclusions

Abbreviations

Declarations

References

Tables

Supplementary Files

Status:

Version 1