RESUMO
Objective:To investigate the characteristics and reliability of a novel sub-classification of Wilkins type III lateral-flexion supracondylar fracture of the humerus in children.Methods:A retrospective cohort study was used to analyze the clinical data of 92 children with supracondylar fracture of the humerus admitted to Provincial Children′s Hospital of Anhui Medical University from January 2013 to August 2021, including 38 males and 54 females, aged 2-13 years [(8.5±2.4)years]. Lateral-flexion Wilkins type III supracondylar humeral fractures were classified into two subtypes according to the fracture features: type IIIA ( n=14), complete fracture with the distal fragment displaced anteriorly and laterally, with no obvious anterior or posterior inclination (<10°) or rotation; type IIIB ( n=78), complete fracture with the distal fragment displaced anteriorly and laterally, with significant anterior or posterior inclination (>10°) or rotation. The incidence and risk ratio of ulnar nerve injury and open reduction were compared between the two subtypes of the fracture. The weighted Kappa method was used to test the inter- and intra-observer agreement of the two new subtypes. Results:Of all, 15 children had ulnar nerve injury, among which 1(6.7%) was type IIIA and 14(93.3%) were type IIIB; while other 77 children had no ulnar nerve injury. The risk of ulnar nerve injury in children with type IIIB was 3-fold higher than that in children with type IIIA ( OR=2.84, 95% CI 0.34- 25.56, P>0.05). The open reduction was performed in 11(73.3%) out of the 15 children with ulnar nerve injury, but in 18(23.4%) out of the 77 children with no ulnar nerve injury. The risk of open reduction in children with ulnar nerve injury was 9-fold higher than that in children without ulnar nerve injury ( OR=9.01, 95% CI 2.28- 33.17, P<0.01). Open reduction was performed in 29 children, among which 2(6.9%) were type IIIA and 27(93.1%) were type IIIB. The risk of open reduction in children with type IIIB was 3-fold higher than that in children with type IIIA ( OR=3.17, 95% CI 0.66-15.24, P>0.05). The intra-observer Kappa value was 0.49±0.09(95% CI 0.31-0.66), indicating a moderate agreement. The inter-observer Kappa value was 0.80±0.06(95% CI 0.68-0.91), indicating a strong or very strong agreement. Conclusions:Wilkins type IIIB lateral-flexion supracondylar fracture of the humerus in children is more likely to be accompanied by ulnar nerve injury and to be opt to open reduction in comparion with type IIIA. The new subtyping has reliable inter-observer and intra-observer consistency, and is able to facilitate the prediction of surgical plans.
RESUMO
Objective To investigate the value of renal CT volumetric texture analysis with machine learning radiomics in assessment of pathological grade of clear cell renal cell carcinoma(ccRCC). Methods Thirty-four biopsy-confirmed ccRCC subjects who had four-phase CT scanning (NC:non-contrast, CM: Corticomedullary, N: Nephrographic, E: Excretory) were collected retrospectively from June 2013 to October 2017 for the study.Non-rigid registration was performed on multi-phase CT images in reference to CM-phase.Each lesion was segmented on CM-phase CT images using our in-house volumetric image analysis platform,"3DQI".A set of fifty-nine volumetric textures,including histogram,gradient,gray level co-occurrence matrix(GLCM),run-length(RL),moments,and shape,was calculated for each segment lesion in each phase as parameters for the training/testing of Random Forest (RF) classifier. Four groups according to pathological Fuhrman grade on a scaleⅠtoⅣ,these tumors were then divided into low(Ⅰ+Ⅱ) and high grade ( Ⅲ + Ⅳ) groups. Feature selection was performed by Boruta algorithm. A 10-fold cross-validation method was applied to validate the RF performance by receiver operating characteristic (ROC) curves analysis to determine the diagnostic accuracy of the model. Results Subjects were divided into four groups by Fuhrman grade on a scaleⅠtoⅣ:3 cases gradeⅠ,19 cases gradeⅡ,8 cases gradeⅢand 4 cases gradeⅣ.In CM-phase,kurtosis and long-run-emphasis(RLE)were selected the most important textures for ccRCC staging among 59 features. The area under curve (AUC) of ROC was 0.88 (79% sensitivity and 82% specificity)by using kurtosis and RLE textures.The mean values of kurtosis and RLE were(-20.00±22.00)×10-2and(3.00±0.40)×10-2for low group,whereas(31.00±32.00)×10-2and(5.00± 0.02)×10-2for high group.Within the mean±SD range of statistics,radiomics can distinguish between low and high grade tumors.In multi-phase analysis,three most important features were selected among 236(59× 4) textures: kurtosis (CM-phase), GLCM homogeneity I (HOMO 1) (E-phase), and GLCM homogeneity 2 (HOMO2)(E-phase).The mean values of HOMO 1(E-phase)and HOMO 2(E-phase)were(19.00±0.03)× 10-2and(11.00±0.02)×10-2for low group,whereas(22.00±0.03)×10-2and(14.00±0.02)×10-2for high group. The AUC was 0.92(93% sensitivity and 87% specificity)by using these three textures. Conclusion This study has demonstrated that renal CT volumetric texture analysis with machine learning radiomics could preoperative accurately perform cancer staging for ccRCC.