Objective To investigate the CT features of inflammatory myofibroblastic tumor in children. Methods Eighteen patients with inflammatory myofibroblastic tumor proven by surgery and pathology were examined with plain and contrast medium enhancement CT scan. Results Of 18 cases,16 had isolated lesions located at lung (n =4), mesentery (n =3), kidney (n =2) and trachea (n = 1 ),left main bronchus ( n = 1 ), right thoracic cavity ( n = 1 ), peritoneum cavity ( n = 1 ), pancreas ( n = 1 ),left thigh ( n = 1 ), prostate ( n = 1 ), superclvicle soft t tissue ( n = 1 ) , bladder ( n = 1 ). The other 2 cases were with multiple lesions on omentum and mesentery, and in intraperitoneal and side of split of right hepatic lobe, respectively. The CT findings of 18 cases included 16 solid mass with calcifications in 3 of them, and 2 solid-cystic mass. After contrast enhancement, moderate or marked homogeneous or heterogeneous enhancement were shown in all the solid parts of tumor on dynamic CT. Mass can compress surround great vessel and tube-like structure. On pathological examination, the tumor was mainly composed of spindleshaped fibrous cells and inflammatory cells, and the immunohistochemically staining for SMA was observed positively. Conclusion CT can provide specific information for diagnosis of inflammatory myofibroblastic tumor, yet definite diagnosis relies on pathology.

Objective To investigate the feasibility to reduce radiation doses on pediatric mutidetector abdominal CT using the adaptive statistical iterative reconstruction technique (ASIR) associated with automated tube current modulation technique(ATCM).Methods Thirty patients underwent abdominal CT with ATCM and the follow-up scan with ATCM cooperated with 40％ ASIR.ATCM was used with agedependent noise index (NI) settings: NI =9 for 0-5 year old and NI =11 for ＞ 5 years old for simple ATCM group,NI =11 for 0-5 year old and NI =15 for ＞5 years old for ATCM cooperated with 40％ ASIR group(AISR group).Two radiologists independently evaluated images for diagnostic quality and image noise with subjectively image quality score and image noise score using a 5-point scale.Interobserver agreement was assessed by Kappa test.The volume CT dose indexes (CTDIvol) for the two groups were recorded.Statistical significance for the CTDIvol value was analyzed by pair-sample t test.Results The average CTDIvol for the ASIR group was (1.38 ± 0.64) mGy,about 60％ lower than (3.56 ± 1.23) mGy for the simple ATCM group,and the CTDIvol of two groups had statistically significant differences.(t =33.483,P ＜ 0.05).The subjective image quality scores for the simple ATCM group were 4.43 ± 0.57 and 4.37 ±0.61,Kappa =0.878,P ＜ 0.01 (ASIR group: 4.70 ± 0.47 and 4.60 ± 0.50,Kappa =0.783,P ＜ 0.01),by two observers.The image noise score for the simple ATCM group were 4.03 ±0.56 and 3.83 ±0.53,Kappa =0.572,P ＜ 0.01 (ASIR group: 4.20 ± 0.48 and 4.10 ± 0.48,Kappa =0.748,P ＜ 0.01),by two observers.All images had acceptable diagnostic image quality.Conclusion Lower radiation dose can be achieved by elevating NI with ASIR in pediatric CT abdominal studies,while maintaining diagnostically acceptable images.

Objective To investigate the CT manifestations and clinical features of thymic lymphoepithelioma-like carcinoma(LELC)in children.Methods The CT findings and clinical features of 5 children with LELC confirmed by pathology were analyzed retrospectively,and relevant literatures were reviewed.Results All the five children showed a large soft tissue mass(a few punctuate calcifications in 3 cases)in the anterior and middle mediastinum,which was lobulated.After enhancement,the tumor showed signifi-cant heterogeneous enhancement with cystic,necrosis,compressing,surrounding and invading adjacent structures.Twenty-one children with LELC were searched in literatures.Conclusion Thymic LELC in children usually occurs in adolescent boys aged 10-16 years.Those children usually go to the doctor with symptoms of tumor compression or invasion of adjacent structures.LELC shows advanced clinical stage with poor prognosis.CT enhancement can accurately display the lesion and invasion of adjacent structures,and guide clinical staging and treatment.

Objective:To identify risk factors associated with mandibular osteoradionecrosis (ORN) in oral and maxillofacial cancer patients following radiotherapty and to provide scientific basis for the etiological research and clinical prevention of mandibular ORN.Methods:A retrospective study was conducted in patients with oral and maxillofacial-head and neck cancer during the period from January 2013 to December 2015. Influential factors related to mandibular ORN were screened by single factor analysis, Lasso and Logistic regression analysis.Results:A total of 757 patients were analyzed, and the total incidence of mandibular ORN was 12.0%(91/757). There were 443 males and 314 females, aged (51.8±13.7) years. Thirty-five related factors were screened to 28 by single factor analysis. It was determined by Lasso regression analysis that, radiation doses ( OR=1.135, P=0.034, 95% CI: 1.089-1.232), T classification ( OR=2.586, P=0.001, 95% CI: 1.482-4.512), mandibular surgery ( OR=9.101, P<0.001, 95% CI: 2.796-29.630), periodontitis ( OR=6.089, P<0.001, 95% CI: 2.708-13.693), diabetes ( OR=4.467, P=0.002, 95% CI: 1.705-11.704), tooth extraction after radiotherapy ( OR=3.228, P=0.001, 95% CI: 1.640-6.350), dental caries ( OR=2.911, P=0.009, 95% CI: 1.300-6.516), periapical periodontitis ( OR=2.726, P=0.016, 95% CI: 1.209-6.145), smoking ( OR=4.438, P=0.002, 95% CI: 1.702-11.571) and unilateral/bilateral radiotherapy ( OR=2.225, P=0.028, 95% CI: 1.090-4.545) were significantly associated with developing mandibular ORN. Conclusions:Ten main risk factors for mandibular ORN were identified through the single center, large sample, retrospective analysis, which has a certain value for clinical prevention of mandibular ORN. Prospective, randomized controlled trials and long-term follow-up are still needed.