Machine learning versus regression for prediction of sporadic pancreatic cancer

被引：16

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 32 条

[1]

Allison P., 1995, SURVIVAL ANAL USING

[2] Individual risk prediction: Comparing random forests with Cox proportional-hazards model by a simulation study [J].

Baralou, Valia ;

Kalpourtzi, Natasa ;

Touloumi, Giota .

BIOMETRICAL JOURNAL, 2023, 65 (06)

[3]

Chen T., XGBOOST EXTREME GRAD

[4] XGBoost: A Scalable Tree Boosting System [J].

Chen, Tianqi ;

Guestrin, Carlos .

KDD'16: PROCEEDINGS OF THE 22ND ACM SIGKDD INTERNATIONAL CONFERENCE ON KNOWLEDGE DISCOVERY AND DATA MINING, 2016, :785-794

[5] Derivation and External Validation of Machine Learning-Based Model for Detection of Pancreatic Cancer [J].

Chen, Wansu ;

Zhou, Yichen ;

Xie, Fagen ;

Butler, Rebecca K. ;

Jeon, Christie Y. ;

Luong, Tiffany Q. ;

Zhou, Botao ;

Lin, Yu-Chen ;

Lustigova, Eva ;

Pisegna, Joseph R. ;

Kim, Sungjin ;

Wu, Bechien U. .

AMERICAN JOURNAL OF GASTROENTEROLOGY, 2023, 118 (01) :157-167

[6]

Chen Wansu, 2019, Perm J, V23, DOI 10.7812/TPP/18-213

[7]

COX DR, 1972, J R STAT SOC B, V34, P187

[8] NOVEL HEAD AND NECK CANCER SURVIVAL ANALYSIS APPROACH: RANDOM SURVIVAL FORESTS VERSUS COX PROPORTIONAL HAZARDS REGRESSION [J].

Datema, Frank R. ;

Moya, Ana ;

Krause, Peter ;

Baeck, Thomas ;

Willmes, Lars ;

Langeveld, Ton ;

de Jong, Robert J. Baatenburg ;

Blom, Henk M. .

HEAD AND NECK-JOURNAL FOR THE SCIENCES AND SPECIALTIES OF THE HEAD AND NECK, 2012, 34 (01) :50-58

[9] The Multicenter Cancer of Pancreas Screening Study: Impact on Stage and Survival [J].

Dbouk, Mohamad ;

Katona, Bryson W. ;

Brand, Randall E. ;

Chak, Amitabh ;

Syngal, Sapna ;

Farrell, James J. ;

Kastrinos, Fay ;

Stoffel, Elena M. ;

Blackford, Amanda L. ;

Rustgi, Anil K. ;

Dudley, Beth ;

Lee, Linda S. ;

Chhoda, Ankit ;

Kwon, Richard ;

Ginsberg, Gregory G. ;

Klein, Alison P. ;

Kamel, Ihab ;

Hruban, Ralph H. ;

He, Jin ;

Shin, Eun Ji ;

Lennon, Anne Marie ;

Canto, Marcia Irene ;

Goggins, Michael .

JOURNAL OF CLINICAL ONCOLOGY, 2022, 40 (28) :3257-+

[10] Tests of calibration and goodness-of-fit in the survival setting [J].

Demler, Olga V. ;

Paynter, Nina P. ;

Cook, Nancy R. .

STATISTICS IN MEDICINE, 2015, 34 (10) :1659-1680

← 1 2 3 4 →