Objective:To evaluate the value of ultrasound radiomics combined with machine learning for early diagnosis of seronegative Hashimoto’s thyroiditis (SN-HT) .Methods:This retrospective study included 164 patients from Liaoning Provincial People’s Hospital , Lixin County People’s Hospital, Linghai Dalinghe Hospital, Fengcheng Phoenix Hospital, who underwent thyroidectomy for solitary nodules with normal thyroid function between Nov. 2016 and Jan. 2024. Postoperative pathology confirmed Hashimoto’s thyroiditis (HT) in some cases, who were further categorized into antibody-positive and antibody-negative groups based on serum antibody status. Patients without Hashimoto’s thyroiditis served as the control group. A total of 298 ultrasound images were analyzed. Radiomics features were extracted from hypoechoic non-nodular areas within 0.5 cm surrounding the tumor. Two senior pathologists and two senior ultrasound physicians independently assessed lymphocytic infiltration, eosinophilic changes of follicular epithelium, and the proportion of hypoechoic areas in pathology and ultrasound images, respectively. A machine learning model, CCH-NET, was developed using linear regression and t-distributed stochastic neighbor embedding (t-SNE) techniques. The dataset was divided into a training set (80%) and a validation set (20%) to compare the diagnostic accuracy of CCH-NET with that of senior ultrasound physicians. Results:In internal validation, CCH-NET achieved a diagnostic accuracy of 88.89% for both antibody-positive and antibody-negative groups, significantly higher than the 66.67% accuracy of senior ultrasound physicians ( P<0.01). In external validation, CCH-NET achieved 75.00% and 66.67% accuracy for the two groups, compared to 50.00% by senior ultrasound physicians. For the control group, both methods achieved 93.33% accuracy. The AUC of CCH-NET was 0.848, outperforming senior ultrasound physicians (0.681) ,demonstrating superior diagnostic performance. Conclusion:The radiomics-based CCH-NET model, using non-nodular hypoechoic areas as a specific indicator, can accurately identify early SN-HT in euthyroid patients. It significantly outperforms senior ultrasound physicians, improving diagnostic accuracy and reducing missed diagnoses.

Objective:To evaluate the value of ultrasound radiomics combined with machine learning for early diagnosis of seronegative Hashimoto’s thyroiditis (SN-HT) .Methods:This retrospective study included 164 patients from Liaoning Provincial People’s Hospital , Lixin County People’s Hospital, Linghai Dalinghe Hospital, Fengcheng Phoenix Hospital, who underwent thyroidectomy for solitary nodules with normal thyroid function between Nov. 2016 and Jan. 2024. Postoperative pathology confirmed Hashimoto’s thyroiditis (HT) in some cases, who were further categorized into antibody-positive and antibody-negative groups based on serum antibody status. Patients without Hashimoto’s thyroiditis served as the control group. A total of 298 ultrasound images were analyzed. Radiomics features were extracted from hypoechoic non-nodular areas within 0.5 cm surrounding the tumor. Two senior pathologists and two senior ultrasound physicians independently assessed lymphocytic infiltration, eosinophilic changes of follicular epithelium, and the proportion of hypoechoic areas in pathology and ultrasound images, respectively. A machine learning model, CCH-NET, was developed using linear regression and t-distributed stochastic neighbor embedding (t-SNE) techniques. The dataset was divided into a training set (80%) and a validation set (20%) to compare the diagnostic accuracy of CCH-NET with that of senior ultrasound physicians. Results:In internal validation, CCH-NET achieved a diagnostic accuracy of 88.89% for both antibody-positive and antibody-negative groups, significantly higher than the 66.67% accuracy of senior ultrasound physicians ( P<0.01). In external validation, CCH-NET achieved 75.00% and 66.67% accuracy for the two groups, compared to 50.00% by senior ultrasound physicians. For the control group, both methods achieved 93.33% accuracy. The AUC of CCH-NET was 0.848, outperforming senior ultrasound physicians (0.681) ,demonstrating superior diagnostic performance. Conclusion:The radiomics-based CCH-NET model, using non-nodular hypoechoic areas as a specific indicator, can accurately identify early SN-HT in euthyroid patients. It significantly outperforms senior ultrasound physicians, improving diagnostic accuracy and reducing missed diagnoses.

ObjectiveTo explore the longitudinal measurement invariance of standardized Constitution in Chinese Medicine Questionnaire （CCMQ） among Chinese adult populations， and to provide evidence for the longitudinal effect analysis of this questionnaire. MethodsA total of 509 adults who voluntarily received 26 weeks of comprehensive traditional Chinese medicine （TCM） constitution fitness intervention were included （3 time points： before intervention， 13 weeks intervention， and 26 weeks intervention）， evaluated by CCMQ with 9 subscales. A single-group confirmatory factor analysis was performed to establish a single-group baseline model with a well-fitting model， and then nested models， that is multiple groups of confirmatory factors analysis， are used to analyze the longitudinal measurement invariance， followed by configural invariance， metric invariance， scalar invariance， and strict invariance. Comparative Fit Index （CFI）， Root Mean Square Error of Approximation （RMSEA）， and Standardized Root Mean Square Residual （SRMR） were used as model fit indicators for the analysis. ResultsThe single-group confirmatory factors analysis proved that the 9 subscales of CCMQ fit well at the three time points （CFI： 0.933~0.992， RMSEA： 0.027~0.080， SRMR： 0.022~0.045）. The multiple-group confirmatory factors analysis showed that the configural， metric， scalar， and strict invariance across time of qi-stagnation and special diathesis constitution were established （ΔCFI＜0.01， ΔRMSEA＜0.01）； the configural， metric， scalar， and partial strict invariance across time of phlegm-dampness constitution subscale were established （ΔCFI＜0.01， ΔRMSEA＜0.01）； the configural， metric， partial scalar and partial strict invariance of 6 subscales， including gentleness， qi-deficiency， yang-deficiency， yin-deficiency， dampness-heat， and blood-stasis constitutions， were all established （ΔCFI＜0.01， ΔRMSEA＜0.01）. ConclusionThe CCMQ met at least metric invariance at 3 time points， which can be applied to the study of whether different time points have the same unit or meaning. The measurement invariance of the 3 subscales of phlegm-dampness， qi-stagnation and special diathesis， special diathesis have longitudinal measurement invariance， so the mean comparison over time could be made； while the 6 subscales of gentleness， qi-deficiency， yang-deficiency， yin-deficiency， dampness-heat， and blood-stasis constitutions meet partial scalar and partial strict invariance， which could be explained partially for the difference in the comparison of the mean over time.

ObjectiveTo construct and evaluate the transformation model of the Constitution in Chinese Medicine Questionnaire （CCMQ） to SF-6Dv1 health utility and broaden the applications of CCMQ. MethodsThe data of CCMQ and SF-6Dv1 were collected from 595 participants at baseline， 3 months and 6 months after the comprehensive intervention suitable for the corresponding traditional Chinese medicine （TCM） constitution. The estimation and validation datasets were constructed， and four statistical algorithms including the ordinary least squares （OLS）， MM robust regression （MM）， censored least absolute deviations （CLAD） and the Tobit model were used to create alternative models. The mean absolute error （MAE）， root mean square error （RMSE） and intraclass correlation coefficient （ICC） were used to evaluate the prediction performance of the model. ResultsThe constitution scores of all TCM constitutions by CCMQ was significantly correlated with the SF-6Dv1 health utility value measured at three timepoints； the health utility value of the SF-6Dv1 was positively correlated with gentleness type （r=0.596， r=0.578， r=0.606， all P＜0.05） and negatively correlated with eight unbalanced constitutions （r=-0.586~-0.301， all P＜0.05）. The MM established based on the subscale scores of CCMQ was the optimal mapping model， and the MAE， RMSE， and ICC values were 0.0741， 0.0930 and 0.766， respectively. Gentleness type， qi-deficiency type， phlegm-wetness type， qi-constraint type， and age were the primary factors included in the model. The measured and predicted value of SF-6Dv1 had a moderate positive correlation （r=0.673， r=0.617， P＜0.05） and a good consistency as shown by the Bland-Altman plot. ConclusionBy using MM， the CCMQ can be transformed into SF-6Dv1 health utility value for health economics analysis.

Objective To evaluate the efficacy and safety of X-ray guided endoscopic gastrojejunostomy using stent in treatment of malignant gastric outlet obstruction ( GOO ) . Methods Six hospitalized patients with malignant GOO underwent X-ray guided endoscopic gastrojejunostomy using stent in the department of gastroenterology, Shandong Provincial Hospital Affiliated to Shandong University between March 2017 and June 2017. The technical success rate, clinical success rate, procedure time, adverse events and follow-up were recorded and analyzed in this retrospective study. Results The stent was successfully placed in the 6 patients with 100％ ( 6/6) technical success rate. The mean procedure time was 91. 7±51. 8 min. After the procedure, all patients were fed liquid or semi-liquid diet, and the GOO score system was increased from 0-1 before operation to 2-3 after operation. The clinical success rate was 100％(6/6). Peritonitis was observed in 2 patients during operation, and resolved by abdominal drainage. Gastrointestinal bleeding occurred in 1 patient after operation, which was resolved with conservative treatment. During a mean follow-up period of 78. 6 days (range 32-100 days), there was no recurrence of obstruction symptoms except that 1 patient died because of tumor progress 60 days after procedure. Conclusion The X-ray guided endoscopic gastrojejunostomy using stent is feasible and safe to treat malignant GOO with a reliable short-term efficacy.

Objective:To investigate the association between catechol-O-methyl transferase (COMT)Vall58Met polymorphism and prepulse inhibition of the auditory startle reflex (PPI) in patients with schizophrenia.Methods:Totally 178 patients with schizophrenia and 190 healthy volunteers were recruited.The auditory startle reflex was detected by using SR-HLAB monitoring system.The indexed of the auditory startle reflex included the amplitude,habituation％ and PPI30,PPI60,PPI120 (the lead interval was set 30 ms,60 ms,120 ms).COMT Vall58Met polymorphism was genotyped by polymerase chain reaction-restriction fragment length polymorphism (PCR-RELP).The differences of PPI among COMT genotypes were compared.Results:Compared to the healthy volunteers,patients with schizophrenia had a significant lower the amplitude of auditory startle reflex[(563± 460) mV vs (695 ± 447) mY,P ＜ 0.05] and habituation％ [(32 ± 46) vs (48 ± 33),P ＜ 0.01] as well as the ％PPI120[(27 ± 5) vs (35 ± 3),P ＜ 0.05].The significant differences in COMT allelic and genotypic distribions were observed between patients with schizophrenia and healthy volunteers (x2 =8.16,11.74,Ps ＜ 0.05).The significant main effect of COMT genotype on habituation％ was observed (P ＜0.05) but no interaction genotype by diagnosis on the amplitude of auditory startle reflex,habituation％ and ％ PPI120 was observed (Ps ＞ 0.05).Conclusions:There may be a correlation between COMT genotype and adaptability,but not between COMT genotype and PPI deficit present in patients with schizophrenia

Objective To investigate the relationship between body mass index (BMI) and health-related quality of life (HRQOL) in the adult Chinese population.Methods Data of 21 108 adults was collected from a cross-sectional survey performed in 9 provinces of China (Jiangsu,Anhui,Gansu,Qinghai,Fujian,Beijing,Jilin,Jiangxi and Henan province).Rank sum test was used to compare HRQOL with different BMI category.Multiple logistic regression analysis was used to assess the impact of different BMI category on HRQOL after adjusted for sex,age,marital,education,physical activity status,and tobacco and alcohol addiction.Results (1) For healthy subjects,the overweight BMI group had the highest HRQOL score in the physical component summary (PCS) (83.69±14.09) and mental component summary (MCS)(80.29± 15.41),while the underweight group had the worst HRQOL(81.20-± 14.05,74.82± 16.41).For subjects with chronic condition,overweight had the highest H RQOL score in the PCS(74.63± 18.51),and obese group had the best HRQOL in the MCS (76.75± 18.83).The variation among healthy and chronic disease subjects was much greater than the differences among BMI category groups.(2) Compared with normal weight,data on odds ratio(ORs) of impaired HRQOL in PCS(healthy subjects OR=1.23(1.10-1.35),chronic disease subjects OR=1.48(1.20-1.82)) MCS(healthy subjects OR=1.14(1.03-1.26),chronic disease subjects OR=1.37 (1.11-1.68)) and multiple dimensions increased among underweight.ORs of impaired HRQOL in PCS and general health dimension,MCS (healthy subjects OR =0.81 (0.72-0.91),chronic disease subjects OR=0.80(0.71-0.91)) and multiple dimensions decreased among overweight.ORs of impaired HRQOL in physical functioning dimension (healthy subjects OR=1.91 (1.25-2.92),chronic disease subjects OR=1.65(1.21-2.26)) while in MCS and role emotional and mental health dimension decreased among obese.Conclusion Whatever health or chronic disease,the influence of BMI on HRQOL is similar:the HRQOL score for the underweight group is significantly lower than that for other BMI groups in PCS and MCS.Overweight and the obese people has better HRQOL in MCS,and the obese people has poor physical function.The relation above between BMI and HRQOL is more obvious in people with chronic disease.

OBJECTIVETo examine the function of body mass index (BMI) as a moderator reflecting the relationship between chronic disease and health-related quality of life (HRQOL).

METHODSThis study included 8 314 participants pooled from a general population-based cross-sectional survey that had been conducted in Beijing and 8 provinces of China (Jiangsu,Anhui,Gansu, Qinghai, Fujian, Jilin, Jiangxi, and Henan). Hierarchical multiple regression was emplayed to test the moderating effect.

RESULTSIn physical component summary of SF-36, the regression coefficient of interaction on chronic disease and BMI was not significantly different (β = 0.084, P = 0.142), while the new ΔR(2) was not significantly different (ΔR(2) = 0.000, P = 0.142) either. In mental component summary of SF-36, the interaction on chronic disease and BMI was significantly different (β = 0.132, P = 0.034), so as the new ΔR(2) (ΔR(2) = 0.001, P = 0.034). Compared to the standard regression coefficient, chronic disease had a greater negative impact on HRQOL than BMI on both physical and mental component summaries.

CONCLUSIONOur results indicated that BMI could moderate the association between chronic disease and HRQOL. The higher the BMI, the smaller negative impact of chronic disease on HRQOL in mental component summary was seen.

Objective: To analyze the application status of the EQ-5D and SF-6D in cost-utility analysis ( CUA) . Methods: The data from PubMed data base from 2003-2012 was conducted as the main resource in econometric analysis. Results: There is no certain increasing trend of literature quantity in the recent 10 years’ application of EQ-5D and SF-6D in the CUA. The main published language was English and the main country was the United Kingdom. The core authors of researches on the CUA estimated by EQ-5D and SF-6D had not been formed, however, the core periodicals had been formed. The literatures involve wide-spread diseases, mainly known as Motion system and pain diseases. Most of the effectiveness could be calculated with quality adjusted life year ( QALY) , the object of CUA is mainly on the medicine, operation and health intervention. Conclusion: Appicating EQ-5D and SF-6D in CUA and research content of the literatures has wide distribution width, although it has shown centralized trend, which still needs more studies by related researchers.

Objective To examine the function of body mass index(BMI)as a moderator reflecting the relationship between chronic disease and health-related quality of life(HRQOL). Methods This study included 8 314 participants pooled from a general population-based cross-sectional survey that had been conducted in Beijing and 8 provinces of China(Jiangsu,Anhui,Gansu, Qinghai,Fujian,Jilin,Jiangxi,and Henan). Hierarchical multiple regression was emplayed to test the moderating effect. Results In physical component summary of SF-36,the regression coefficient of interaction on chronic disease and BMI was not significantly different(β=0.084,P=0.142),while the new ΔR2 was not significantly different (ΔR2=0.000,P=0.142) either. In mental component summary of SF-36,the interaction on chronic disease and BMI was significantly different(β=0.132, P=0.034),so as the new ΔR2(ΔR2=0.001,P=0.034). Compared to the standard regression coefficient,chronic disease had a greater negative impact on HRQOL than BMI on both physical and mental component summaries. Conclusion Our results indicated that BMI could moderate the association between chronic disease and HRQOL. The higher the BMI,the smaller negative impact of chronic disease on HRQOL in mental component summary was seen.