Objective To explore the value of machine learning(ML)models based on ultrasonic texture features(TF)of coronary artery for predicting incomplete Kawasaki disease(IKD)in children.Methods Forty-eight children with IKD and 48 children without KD(non-KD)were enrolled with propensity score matching and divided into training set(n=67,34 cases of IKD and 33 cases of non-KD)and test set(n=29,14 of IKD and 15 of non-KD)at the ratio of 7∶3.Based on clinic-laboratory indicators(C-L)in training set and TF obtained with texture analysis of coronary artery ultrasound images,the optimal C-L-related features and TF were selected.Based on the optimal C-L correlated features,TF and their combinations,6 ML models,including random forest(RF),support vector machine(SVM),logistic regression(LR),gradient boosting decision tree(GBDT),decision tree(DT)and eXtreme gradient boosting(XGBoost)were respectively constructed for predicting IKD in children.The models were then trained in training set and validated in test set,and the best C-L ML,TF ML and C-L-TF ML models were selected.The area under the curve(AUC)of the best ML models were compared,and the clinical value of the best TF ML model was observed with decision curve analysis(DCA).Results Totally 3 optimal C-L related features and 8 optimal TF were selected.Among the constructed C-L ML,TF ML and C-L-TF ML models,C-L-LR model,TF-LR model and C-L-TF-SVM model were the optimal ones,with AUC in training set of 0.891,0.985 and 0.965,while in test set of 0.676,0.971 and 0.948,respectively.No significant difference of AUC was found between TF-LR model and C-L-TF-SVM model in both training set and test set(both P＞0.05),which were both greater than those of C-L-LR model(all P＜0.05).TF-LR model achieved higher clinical benefits in both training set and test set.Conclusion Ultrasound TF-LR model of coronary artery could be used to effectively predict IKD in children.

Objective To explore the value of machine learning(ML)models based on ultrasonic texture features(TF)of coronary artery for predicting incomplete Kawasaki disease(IKD)in children.Methods Forty-eight children with IKD and 48 children without KD(non-KD)were enrolled with propensity score matching and divided into training set(n=67,34 cases of IKD and 33 cases of non-KD)and test set(n=29,14 of IKD and 15 of non-KD)at the ratio of 7∶3.Based on clinic-laboratory indicators(C-L)in training set and TF obtained with texture analysis of coronary artery ultrasound images,the optimal C-L-related features and TF were selected.Based on the optimal C-L correlated features,TF and their combinations,6 ML models,including random forest(RF),support vector machine(SVM),logistic regression(LR),gradient boosting decision tree(GBDT),decision tree(DT)and eXtreme gradient boosting(XGBoost)were respectively constructed for predicting IKD in children.The models were then trained in training set and validated in test set,and the best C-L ML,TF ML and C-L-TF ML models were selected.The area under the curve(AUC)of the best ML models were compared,and the clinical value of the best TF ML model was observed with decision curve analysis(DCA).Results Totally 3 optimal C-L related features and 8 optimal TF were selected.Among the constructed C-L ML,TF ML and C-L-TF ML models,C-L-LR model,TF-LR model and C-L-TF-SVM model were the optimal ones,with AUC in training set of 0.891,0.985 and 0.965,while in test set of 0.676,0.971 and 0.948,respectively.No significant difference of AUC was found between TF-LR model and C-L-TF-SVM model in both training set and test set(both P＞0.05),which were both greater than those of C-L-LR model(all P＜0.05).TF-LR model achieved higher clinical benefits in both training set and test set.Conclusion Ultrasound TF-LR model of coronary artery could be used to effectively predict IKD in children.

Objective:To compare the difference of LLD (leg length discrepancy) between robot-assisted and conventional methods of total hip arthroplasty (THA).Methods:Data of 38 patients who had THA performed by robot-assisted or conventional methods from January 2019 to May 2020 were retrospectively analyzed. There were 38 cases (54 hips) in robot-assisted THA group (robot group) with 18 males and 20 females (age 53.5±13.6 years, BMI 26.2±3.4 kg/m 2), and there were 21 cases (32 hips) with osteonecrosis of the femoral head, 17 cases (22 hips) with Crown typeⅠandⅡdevelopmental dysplasia of the hip. There were 38 cases (54 hips) in conventional THA group (conventional group), with 19 males and 19 females, (age 52.3±14.7 years old, BMI 25.7±2.9 kg/m 2), and there were 19 cases (30 hips) with developmental dysplasia of the hip, and 19 cases (24 hips) with osteonecrosis of the femoral head. The operative time, postoperative LLD, Harris score, forgotten joint score-12 (FJS-12) and the difference between preoperative and postoperative LLD between the two groups were compared, and the correlation between surgical methods and the change of hip length was also evaluated. Results:The operation time of the robot group was 73.3±14.1 min and which was 59.3±12.6 min in conventional THA group ( t=2.732, P=0.003). In the robot group, the postoperative LLD was 2.3±3.4 mm, which was less than that of the conventional group 6.7±5.4 mm ( t=3.521, P < 0.001). When the absolute value of LLD was larger than 5 mm as an abnormal value, it was 2.6% (1/38) in the robot group and 47.3% (18/38) in the conventional group. The difference of hip length (HL) in planning and post-operation in the robot group was 2.8±2.2 mm, which was smaller than that in the conventional THA group 7.9±5.3 mm ( t=2.357, P < 0.001). In addition, there was a correlation between the change of hip length results and the postoperative measurement of hip length in the robot group ( r=0.983, P < 0.001). At the last follow-up, Harris score and FJS-12 were recorded in the robot group and coventional group. The scores were 83.1±5.3 and 32.5±4.9 respectively in the robot group, 82.9±7.2 and 31.9±6.7 in the conventional group, respectively. There was no significant difference between the two groups ( t=0.221, 0.356; P=0.819, 0.731). Postoperative bleeding occurred in 1 case in the robot group with postoperative suture healed well. The fracture of the posterior wall of the acetabulum was found in the conventional group and the patient avoids weight bearing 4 weeks after operation. The postoperative recovery was good and no other related complications were found. Conclusion:Robot-assisted THA can accurately restore the length of both legs and reduce LLD compared with conventional THA. The real-time monitoring of LLD during robot operation can give the operator an accurate reference.