Machine learning versus regression for prediction of sporadic pancreatic cancer

被引：12

作者：

Chen, Wansu ^{[1
,6
]}

Zhou, Botao ^{[1
]}

Jeon, Christie Y. ^{[2
]}

Xie, Fagen ^{[1
]}

Lin, Yu-Chen ^{[2
]}

Butler, Rebecca K. ^{[1
]}

Zhou, Yichen ^{[1
]}

Luong, Tiffany Q. ^{[1
]}

Lustigova, Eva ^{[1
]}

Pisegna, Joseph R. ^{[3
,4
]}

Wu, Bechien U. ^{[5
]}

机构：

[1] Kaiser Permanente Southern Calif Res & Evaluat, Pasadena, CA USA

[2] Cedars Sinai Med Ctr, Los Angeles, CA USA

[3] VA Greater Los Angeles Healthcare Syst, Div Gastroenterol & Hepatol, Los Angeles, CA USA

[4] UCLA, David Geffen Sch Med, Dept Med & Human Genet, Los Angeles, CA USA

[5] Southern Calif Permanente Med Grp, Ctr Pancreat Care, Los Angeles Med Ctr, Dept Gastroenterol, Los Angeles, CA USA

[6] Kaiser Permanente Southern Calif, Dept Res & Evaluat, 100 S Robles,2nd Floor, Pasadena, CA 91101 USA

来源：

PANCREATOLOGY | 2023年 / 23卷 / 04期

基金：

美国国家卫生研究院;

关键词：

Risk prediction; Pancreatic cancer; Machine learning versus regression; Random survival forest; eXtreme gradient boosting; RANDOM SURVIVAL FORESTS; MODELS;

D O I：

10.1016/j.pan.2023.04.009

中图分类号：

R57 [消化系及腹部疾病];

学科分类号：

摘要：

Background/objectives: There is currently no widely accepted approach to identify patients at increased risk for sporadic pancreatic cancer (PC). We aimed to compare the performance of two machine-learning models with a regression-based model in predicting pancreatic ductal adenocarcinoma (PDAC), the most common form of PC. Methods: This retrospective cohort study consisted of patients 50-84 years of age enrolled in either Kaiser Permanente Southern California (KPSC, model training, internal validation) or the Veterans Affairs (VA, external testing) between 2008 and 2017. The performance of random survival forests (RSF) and eXtreme gradient boosting (XGB) models were compared to that of COX proportional hazards regression (COX). Heterogeneity of the three models were assessed. Results: The KPSC and the VA cohorts consisted of 1.8 and 2.7 million patients with 1792 and 4582 incident PDAC cases within 18 months, respectively. Predictors selected into all three models included age, abdominal pain, weight change, and glycated hemoglobin (A1c). Additionally, RSF selected change in alanine transaminase (ALT), whereas the XGB and COX selected the rate of change in ALT. The COX model appeared to have lower AUC (KPSC: 0.737, 95% CI 0.710-0.764; VA: 0.706, 0.699-0.714), compared to those of RSF (KPSC: 0.767, 0.744-0.791; VA: 0.731, 0.724-0.739) and XGB (KPSC: 0.779, 0.755-0.802; VA: 0.742, 0.735-0.750). Among patients with top 5% predicted risk from all three models (N 1/4 29,663), 117 developed PDAC, of which RSF, XGB and COX captured 84 (9 unique), 87 (4 unique), 87 (19 unique) cases, respectively. Conclusions: The three models complement each other, but each has unique contributions. & COPY; 2023 IAP and EPC. Published by Elsevier B.V. All rights reserved.

引用

页码：396 / 402

页数：7

共 50 条

[1] Machine Learning Versus Regression for Prediction of Sporadic Pancreatic Cancer
不详
PANCREAS, 2022, 51 (07) : 902 - 902
[2] Improved survival prediction for pancreatic cancer using machine learning and regression
Floyd, Stuart H.
Alvarez, Sergio A.
Ruiz, Carolina
Hayward, John
Sullivan, Mary
Tseng, Jennifer F.
Whalen, Giles F.
GASTROENTEROLOGY, 2007, 132 (04) : A869 - A870
[3] Persistent challenges in pancreatic surgery: Postoperative pancreatic fistula prediction in the machine learning era-Response to: Machine learning versus logistic regression for the prediction of complications after pancreaticoduodenectomy
Muaddi, Hala
Salehinejad, Hojjat
Thiels, Cornelius A.
SURGERY, 2024, 175 (05) : 1466 - 1467
[4] Machine learning versus logistic regression for the prediction of complications after pancreatoduodenectomy
Ingwersen, Erik W.
Stam, Wessel T.
Meijs, Bono J. V.
Roor, Joran
Besselink, Marc G.
Koerkamp, Bas Groot
de Hingh, Ignace H. J. T.
van Santvoort, Hjalmar C.
Stommel, Martijn W. J.
Daams, Freek
SURGERY, 2023, 174 (03) : 435 - 440
[5] Use of Machine Learning in Prediction of Completion of Treatment for Pancreatic Cancer
Khalid, Abdullah
Pasha, Shamsher
Demyan, Lyudmyla
Hartman, Sarah
Levy, Todd
Zanos, Theodoros
Newman, Elliot
Melis, Marcovalerio
ANNALS OF SURGICAL ONCOLOGY, 2024, 31 (01) : S195 - S196
[6] Machine learning versus logistic regression for the prediction of complication after pancreatoduodenectomy Response
Ingwersen, Erik W.
Daams, F.
SURGERY, 2024, 175 (05) : 1467 - 1467
[7] Multimodal survival prediction in advanced pancreatic cancer using machine learning
Keyl, J.
Kasper, S.
Wiesweg, M.
Goetze, J.
Schoenrock, M.
Sinn, M.
Berger, A.
Nasca, E.
Kostbade, K.
Schumacher, B.
Markus, P.
Albers, D.
Treckmann, J.
Schmid, K. W.
Schildhaus, H-U
Siveke, J. T.
Schuler, M.
Kleesiek, J.
ESMO OPEN, 2022, 7 (05)
[8] Invited commentary: Machine learning versus logistic regression for the prediction of complications after pancreatoduodenectomy
Kambakamba, Patryk
SURGERY, 2023, 174 (03) : 441 - 441
[9] Machine learning versus linear regression modelling approach for accurate ozone concentrations prediction
Jumin, Ellysia
Zaini, Nuratiah
Ahmed, Ali Najah
Abdullah, Samsuri
Ismail, Marzuki
Sherif, Mohsen
Sefelnasr, Ahmed
EI-Shafie, Ahmed
ENGINEERING APPLICATIONS OF COMPUTATIONAL FLUID MECHANICS, 2020, 14 (01) : 713 - 725
[10] Risk prediction with machine learning and regression methods
Steyerberg, Ewout W.
van der Ploeg, Tjeerd
Van Calster, Ben
BIOMETRICAL JOURNAL, 2014, 56 (04) : 601 - 606

← 1 2 3 4 5 →