| Backgrounds and Objective:Liver tumors,especially hepatoblastomas,are considered to be one of the most common tumors that occur in children and seriously threaten the lives and health of children.The main method of radical treatment of liver tumors is surgical resection,and insufficient residual liver volume after surgery can cause liver-related postoperative complications.Therefore,it is necessary to analyze the liver volume of the patient before partial hepatectomy to assess whether the residual liver is sufficient to maintain normal liver function.In children,the liver volume changes with growth and development,and the liver volume of children of different ages varies greatly.At the same time,the ratio of liver tumor volume in children to normal liver tissue is often much greater than that of adult liver tumors in normal liver tissue.Therefore,for children with huge liver tumors,enough residual normal liver tissue after liver tumor resection is one of the keys of successful surgical treatment.Across the world,studies on children’s liver volume and children’s standard liver volume formulae are relatively rare.In order to make accurate preoperative planning,it is necessary to estimate the normal liver volume of children.Therefore,the difference in normal liver volume of children of different gender,age,body height and body weight,and the accurate calculation formula of standard liver volume for children to evaluate the liver after surgery are to be explored.This study aims at exploring the growth pattern of liver volume and standard liver volume formulae in children under 14 years of age.Methods:We searched pediatric patients(<14 years of age)with normal liver from the picture archiving and communication system in the Affiliated Hospital of Qingdao University and the Qingdao Women and Children’s Hospital.The study enrolled in 709children(0-14 years)from the Affiliated Hospital of Qingdao University and 59 children(0-14 years)from the Qingdao Women and Children’s Hospital with basic information and CT scan images.Upper abdominal computed tomography(CT)image,age,body height and body weight of the patients were measured and collected before the CT scans.Upper abdominal CT images were reconstructed into three-dimensional model using CT images and Hisense computer assistant system(CAS).Liver volume was automatically measured by CAS.Correlations between ages,body heights,body weights and liver volumes were analyzed by SPSS.New standard liver volume formula was calculated by multivariate linear regression.The significance test of all statistical analysis is two-side test,and the significance level isα=0.05.Results:The liver volumes of the 709 children from the Affiliated Hospital of Qingdao University are ranged from 129.12±14.48ml(<1 month of age,male),141±22.76ml(<1 month of age,female)to 1155.92±155.16ml(male)and1004.17±98.6ml(female).Age,body height and body weight are positively correlated to liver volume.The correlation coefficients of liver volume to age,body height and body weight were 0.90,0.92 and 0.95(P<0.05).New standard liver volume formula derived from body weight,body height and sex was:<20 kg:SLV=3.02×BH+13.3×BW-70.06(r~2=0.82);≥20kg,male:SLV=3.08×BH+12.99BW-72.68(r~2=0.93;BH=body height[cm],BW=body weight[kg]),female:SLV=2.48×BH+13×BW-27.65(r~2=0.93).Conclusion:The study summarized the growth pattern of liver volume and new standard liver volume in children under 14 years of age.Backgrounds and Objective: Gastrointestinal subepithelial lesions(SELs)are tumors originating from layers(including muscularis mucosa,submucosa,or muscularis propria).The two most common SELs are gastrointestinal stromal tumors(GISTs)and gastrointestinal leiomyomas(GILs).All GISTs have malignant potential;however,GILs are considered benign.Endoscopic ultrasonography(EUS),computer tomography,and magnetic resonance imaging cannot effectively distinguish GISTs from GILs.We aimed to develop an artificial intelligence(AI)-based system to differentiate GISTs from GILs using EUS.Methods: The AI system was based on 10439 EUS images of 752 partcipants with histologically-confirmed GISTs or GILs.Participants from different centers were collected to develop and retrospectively evaluate an AI-based system.The system was used to classify GISTs and GILs,when endosonographers considered SELs as GISTs or GILs.We used the system in a multicenter prospective diagnostic test to clinically explore whether joint diagnoses by endosonographers and the AI system can distinguish between GISTs and GILs to improve the total diagnostic accuracy of SELs.Results: The AI system was developed using 10439 EUS images from 752 participants with GISTs or GILs.In the prospective test,132 participants were histologically-diagnosed(36 GISTs,44 GILs,and 52 other types of SELs)among 508 consecutive subjects.Using the AI system,the total accuracy of endosonographers in diagnosing the 132 histologically-confirmed participants with SELs increased from 69.7%(95% confidence interval [CI]: 61.4–76.9%)to 78.8%(95% CI: 71–84.9%;P =0.012).The accuracy of endosonographers in diagnosing the 80 participants with GISTs or GILs increased from 73.8%(95% CI: 63.1–82.2%)to 88.8%(95% CI: 79.8–94.2%;P =0.012).Conclusion: We developed an AI-based system using EUS diagnostic that can help endosonographers effectively distinguish GISTs from GILs and improve the pre-operative diagnostic accuracy of different types of SELs. |
PDF Full Text Request