METHODS: A total of 82 patients(155 eyes)with TAO admitted to the hospital from January 2017 to December 2019 were selected as TAO group. Another 50 patients(100 eyes)with Graves disease without TAO who were admitted during the same period were selected as control group. According to the clinical activity score(CAS)the patients with TAO were divided into active period and inactive period. All the participants were examined by magnetic resonance(MR). The difference of RESOLVE-DWI apparent diffusion coefficient(ADC)values between TAO Group and control group was compared. The receiver operating characteristic curve(ROC)was used to evaluate the diagnostic value of the ADC measurement of RESOLVE-DWI for TAO. Count the related factors that may affect the activity of TAO, and use Logistic regression analysis to clarify the risk factors. ROC curve was used to evaluated the predictive value of RESOLVE-DWI ADC for TAO activity.

RESULTS: The ADC value of the TAO group was greater than that of the control group(P<0.001). ROC curve showed that the best cut-off point of RESOLVE-DWI ADC value in diagnosing TAO was 1.302×10^-3mm²/s, and the sensitivity, specificity and AUC were 87.10%, 81.94% and 0.895, respectively. The age, grading standard promulgated by the American Thyroid Society(NOSPECS)and ADC value of TAO in active period were higher than those in inactive period, and the differences were statistically significant(P<0.05). Logistic regression analysis showed that age, NOSPECS grade and ADC value were the risk factors of TAO activity, and the differences were statistically significant(P<0.05). ROC curve analysis showed that the best cut-off point of RESOLVE-DWI ADC value in predicting TAO activity was 1.522×10^-3mm²/s, and the sensitivity, specificity, and AUC were 82.58%, 76.77%, 0.801, respectively.

CONCLUSION: The best cut-off points of RESOLVE-DWI ADC value in diagnosing TAO and predicting activity are 1.302×10^-3mm²/s and 1.522×10^-3mm²/s, respectively, and use the best cut-off point for TAO diagnosis and activity sexual prediction has high clinical value.