V. A. Suvorov, S. Panin, N. V. Kovalenko, V. Zhavoronkova, M. Postolov, S. Tolstopyatov, A. E. Bublikov, A. V. Panova, V. O. Popova
{"title":"利用机器学习预测胰十二指肠切除术后的胰腺囊肿","authors":"V. A. Suvorov, S. Panin, N. V. Kovalenko, V. Zhavoronkova, M. Postolov, S. Tolstopyatov, A. E. Bublikov, A. V. Panova, V. O. Popova","doi":"10.21294/1814-4861-2023-22-6-25-34","DOIUrl":null,"url":null,"abstract":"Objective: to analyze the results of pancreatoduodenectomy (PD) and identify predictive risk factors for postoperative pancreatic fistula (PF) using machine learning (ML) technology.Material and Methods. A nonrandomized study of treatment outcomes in 128 patients, who underwent PD for periampullary carcinoma between 2018 and 2023, was conducted. To predict PF, the ML models based on the multilayer perceptron and binary logistic regression (BLR) in SPSS Statistics v.26, were used. The Receiver Operator Characteristics (ROC) analysis was used to assess the accuracy of the models. To compare ROC curves, the DeLong test was used.Results. Clinically significant PF occurred in 19 (14.8 %) patients (grade B according to ISGPS 2016 – in 16 (12.5 %), grade C – in 3 (2.3 %)). The data of 90 (70.3 %) patients were used to train the neural network, and 38 (29.7 %) were used to test the predictive model. In multivariate analysis, the predictors of PF were a comorbidity level above 7 points on the age-adjusted Charlson scale, a diameter of the main pancreatic duct less than 3 mm, and a soft pancreatic consistency. The diagnostic accuracy of the ML model estimated using the area under the ROC curve was 0.939 ± 0.027 (95 % CI: 0.859–0.998, sensitivity: 84.2 %, specificity; 96.3 %). The predictive model, which was developed using BLR, demonstrated lower accuracy: 0.918±0.039 (95 % CI: 0.842–0.994, sensitivity: 78.9 %, specificity: 94.5 %) (p=0.02).Conclusion. The use of machine learning technologies makes it possible to increase the probability of a correct prediction of the occurrence of pancreatic fistula after pancreatoduodenectomy.","PeriodicalId":21881,"journal":{"name":"Siberian journal of oncology","volume":" 1","pages":""},"PeriodicalIF":0.0000,"publicationDate":"2024-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Prediction of pancreatic fstula after pancreatoduodenectomy using machine learning\",\"authors\":\"V. A. Suvorov, S. Panin, N. V. Kovalenko, V. Zhavoronkova, M. Postolov, S. Tolstopyatov, A. E. Bublikov, A. V. Panova, V. O. Popova\",\"doi\":\"10.21294/1814-4861-2023-22-6-25-34\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Objective: to analyze the results of pancreatoduodenectomy (PD) and identify predictive risk factors for postoperative pancreatic fistula (PF) using machine learning (ML) technology.Material and Methods. A nonrandomized study of treatment outcomes in 128 patients, who underwent PD for periampullary carcinoma between 2018 and 2023, was conducted. To predict PF, the ML models based on the multilayer perceptron and binary logistic regression (BLR) in SPSS Statistics v.26, were used. The Receiver Operator Characteristics (ROC) analysis was used to assess the accuracy of the models. To compare ROC curves, the DeLong test was used.Results. Clinically significant PF occurred in 19 (14.8 %) patients (grade B according to ISGPS 2016 – in 16 (12.5 %), grade C – in 3 (2.3 %)). The data of 90 (70.3 %) patients were used to train the neural network, and 38 (29.7 %) were used to test the predictive model. In multivariate analysis, the predictors of PF were a comorbidity level above 7 points on the age-adjusted Charlson scale, a diameter of the main pancreatic duct less than 3 mm, and a soft pancreatic consistency. The diagnostic accuracy of the ML model estimated using the area under the ROC curve was 0.939 ± 0.027 (95 % CI: 0.859–0.998, sensitivity: 84.2 %, specificity; 96.3 %). The predictive model, which was developed using BLR, demonstrated lower accuracy: 0.918±0.039 (95 % CI: 0.842–0.994, sensitivity: 78.9 %, specificity: 94.5 %) (p=0.02).Conclusion. The use of machine learning technologies makes it possible to increase the probability of a correct prediction of the occurrence of pancreatic fistula after pancreatoduodenectomy.\",\"PeriodicalId\":21881,\"journal\":{\"name\":\"Siberian journal of oncology\",\"volume\":\" 1\",\"pages\":\"\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2024-01-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Siberian journal of oncology\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.21294/1814-4861-2023-22-6-25-34\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q4\",\"JCRName\":\"Medicine\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Siberian journal of oncology","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.21294/1814-4861-2023-22-6-25-34","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q4","JCRName":"Medicine","Score":null,"Total":0}
Prediction of pancreatic fstula after pancreatoduodenectomy using machine learning
Objective: to analyze the results of pancreatoduodenectomy (PD) and identify predictive risk factors for postoperative pancreatic fistula (PF) using machine learning (ML) technology.Material and Methods. A nonrandomized study of treatment outcomes in 128 patients, who underwent PD for periampullary carcinoma between 2018 and 2023, was conducted. To predict PF, the ML models based on the multilayer perceptron and binary logistic regression (BLR) in SPSS Statistics v.26, were used. The Receiver Operator Characteristics (ROC) analysis was used to assess the accuracy of the models. To compare ROC curves, the DeLong test was used.Results. Clinically significant PF occurred in 19 (14.8 %) patients (grade B according to ISGPS 2016 – in 16 (12.5 %), grade C – in 3 (2.3 %)). The data of 90 (70.3 %) patients were used to train the neural network, and 38 (29.7 %) were used to test the predictive model. In multivariate analysis, the predictors of PF were a comorbidity level above 7 points on the age-adjusted Charlson scale, a diameter of the main pancreatic duct less than 3 mm, and a soft pancreatic consistency. The diagnostic accuracy of the ML model estimated using the area under the ROC curve was 0.939 ± 0.027 (95 % CI: 0.859–0.998, sensitivity: 84.2 %, specificity; 96.3 %). The predictive model, which was developed using BLR, demonstrated lower accuracy: 0.918±0.039 (95 % CI: 0.842–0.994, sensitivity: 78.9 %, specificity: 94.5 %) (p=0.02).Conclusion. The use of machine learning technologies makes it possible to increase the probability of a correct prediction of the occurrence of pancreatic fistula after pancreatoduodenectomy.
期刊介绍:
The main objectives of the journal are: -to promote the establishment of Russia’s leading worldwide positions in the field of experimental and clinical oncology- to create the international discussion platform intended to cover all aspects of basic and clinical cancer research, including carcinogenesis, molecular biology, epidemiology, cancer prevention, diagnosis and multimodality treatment (surgery, chemotherapy, radiation therapy, hormone therapy), anesthetic management, medical and social rehabilitation, palliative care as well as the improvement of life quality of cancer patients- to encourage promising young scientists to be actively involved in cancer research programs- to provide a platform for researches and doctors all over the world to promote, share, and discuss various new issues and developments in cancer related problems. (to create a communication platform for the expansion of cooperation between Russian and foreign professional associations).- to provide the information about the latest worldwide achievements in different fields of oncology The most important tasks of the journal are: -to encourage scientists to publish their research results- to offer a forum for active discussion on topics of major interest - to invite the most prominent Russian and foreign authors to share their latest research findings with cancer research community- to promote the exchange of research information, clinical experience, current trends and the recent developments in the field of oncology as well as to review interesting cases encountered by colleagues all over the world- to expand the editorial board and reviewers with the involvement of well-known Russian and foreign experts- to provide open access to full text articles- to include the journal into the international database- to increase the journal’s impact factor- to promote the journal to the International and Russian markets