Machine learning to predict the occurrence of bisphosphonate-related osteonecrosis of the jaw associated with dental extraction: A preliminary report

Dong Wook Kim, Hwiyoung Kim, Woong Nam, Hyung Jun Kim, In Ho Cha

Research output: Contribution to journalArticle

8 Citations (Scopus)

Abstract

Introduction: The aim of this study was to build and validate five types of machine learning models that can predict the occurrence of BRONJ associated with dental extraction in patients taking bisphosphonates for the management of osteoporosis. Patients & methods: A retrospective review of the medical records was conducted to obtain cases and controls for the study. Total 125 patients consisting of 41 cases and 84 controls were selected for the study. Five machine learning prediction algorithms including multivariable logistic regression model, decision tree, support vector machine, artificial neural network, and random forest were implemented. The outputs of these models were compared with each other and also with conventional methods, such as serum CTX level. Area under the receiver operating characteristic (ROC) curve (AUC) was used to compare the results. Results: The performance of machine learning models was significantly superior to conventional statistical methods and single predictors. The random forest model yielded the best performance (AUC = 0.973), followed by artificial neural network (AUC = 0.915), support vector machine (AUC = 0.882), logistic regression (AUC = 0.844), decision tree (AUC = 0.821), drug holiday alone (AUC = 0.810), and CTX level alone (AUC = 0.630). Conclusions: Machine learning methods showed superior performance in predicting BRONJ associated with dental extraction compared to conventional statistical methods using drug holiday and serum CTX level. Machine learning can thus be applied in a wide range of clinical studies.

Original languageEnglish
Pages (from-to)207-214
Number of pages8
JournalBone
Volume116
DOIs
Publication statusPublished - 2018 Nov

Fingerprint

Bisphosphonate-Associated Osteonecrosis of the Jaw
Tooth Extraction
Area Under Curve
Decision Trees
Holidays
Logistic Models
Machine Learning
Diphosphonates
Serum
ROC Curve
Pharmaceutical Preparations
Osteoporosis
Medical Records
Case-Control Studies

All Science Journal Classification (ASJC) codes

  • Endocrinology, Diabetes and Metabolism
  • Physiology
  • Histology

Cite this

@article{3e1c126a0e3641fc81f516c4ceb28a08,
title = "Machine learning to predict the occurrence of bisphosphonate-related osteonecrosis of the jaw associated with dental extraction: A preliminary report",
abstract = "Introduction: The aim of this study was to build and validate five types of machine learning models that can predict the occurrence of BRONJ associated with dental extraction in patients taking bisphosphonates for the management of osteoporosis. Patients & methods: A retrospective review of the medical records was conducted to obtain cases and controls for the study. Total 125 patients consisting of 41 cases and 84 controls were selected for the study. Five machine learning prediction algorithms including multivariable logistic regression model, decision tree, support vector machine, artificial neural network, and random forest were implemented. The outputs of these models were compared with each other and also with conventional methods, such as serum CTX level. Area under the receiver operating characteristic (ROC) curve (AUC) was used to compare the results. Results: The performance of machine learning models was significantly superior to conventional statistical methods and single predictors. The random forest model yielded the best performance (AUC = 0.973), followed by artificial neural network (AUC = 0.915), support vector machine (AUC = 0.882), logistic regression (AUC = 0.844), decision tree (AUC = 0.821), drug holiday alone (AUC = 0.810), and CTX level alone (AUC = 0.630). Conclusions: Machine learning methods showed superior performance in predicting BRONJ associated with dental extraction compared to conventional statistical methods using drug holiday and serum CTX level. Machine learning can thus be applied in a wide range of clinical studies.",
author = "Kim, {Dong Wook} and Hwiyoung Kim and Woong Nam and Kim, {Hyung Jun} and Cha, {In Ho}",
year = "2018",
month = "11",
doi = "10.1016/j.bone.2018.04.020",
language = "English",
volume = "116",
pages = "207--214",
journal = "Bone",
issn = "8756-3282",
publisher = "Elsevier Inc.",

}

Machine learning to predict the occurrence of bisphosphonate-related osteonecrosis of the jaw associated with dental extraction : A preliminary report. / Kim, Dong Wook; Kim, Hwiyoung; Nam, Woong; Kim, Hyung Jun; Cha, In Ho.

In: Bone, Vol. 116, 11.2018, p. 207-214.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Machine learning to predict the occurrence of bisphosphonate-related osteonecrosis of the jaw associated with dental extraction

T2 - A preliminary report

AU - Kim, Dong Wook

AU - Kim, Hwiyoung

AU - Nam, Woong

AU - Kim, Hyung Jun

AU - Cha, In Ho

PY - 2018/11

Y1 - 2018/11

N2 - Introduction: The aim of this study was to build and validate five types of machine learning models that can predict the occurrence of BRONJ associated with dental extraction in patients taking bisphosphonates for the management of osteoporosis. Patients & methods: A retrospective review of the medical records was conducted to obtain cases and controls for the study. Total 125 patients consisting of 41 cases and 84 controls were selected for the study. Five machine learning prediction algorithms including multivariable logistic regression model, decision tree, support vector machine, artificial neural network, and random forest were implemented. The outputs of these models were compared with each other and also with conventional methods, such as serum CTX level. Area under the receiver operating characteristic (ROC) curve (AUC) was used to compare the results. Results: The performance of machine learning models was significantly superior to conventional statistical methods and single predictors. The random forest model yielded the best performance (AUC = 0.973), followed by artificial neural network (AUC = 0.915), support vector machine (AUC = 0.882), logistic regression (AUC = 0.844), decision tree (AUC = 0.821), drug holiday alone (AUC = 0.810), and CTX level alone (AUC = 0.630). Conclusions: Machine learning methods showed superior performance in predicting BRONJ associated with dental extraction compared to conventional statistical methods using drug holiday and serum CTX level. Machine learning can thus be applied in a wide range of clinical studies.

AB - Introduction: The aim of this study was to build and validate five types of machine learning models that can predict the occurrence of BRONJ associated with dental extraction in patients taking bisphosphonates for the management of osteoporosis. Patients & methods: A retrospective review of the medical records was conducted to obtain cases and controls for the study. Total 125 patients consisting of 41 cases and 84 controls were selected for the study. Five machine learning prediction algorithms including multivariable logistic regression model, decision tree, support vector machine, artificial neural network, and random forest were implemented. The outputs of these models were compared with each other and also with conventional methods, such as serum CTX level. Area under the receiver operating characteristic (ROC) curve (AUC) was used to compare the results. Results: The performance of machine learning models was significantly superior to conventional statistical methods and single predictors. The random forest model yielded the best performance (AUC = 0.973), followed by artificial neural network (AUC = 0.915), support vector machine (AUC = 0.882), logistic regression (AUC = 0.844), decision tree (AUC = 0.821), drug holiday alone (AUC = 0.810), and CTX level alone (AUC = 0.630). Conclusions: Machine learning methods showed superior performance in predicting BRONJ associated with dental extraction compared to conventional statistical methods using drug holiday and serum CTX level. Machine learning can thus be applied in a wide range of clinical studies.

UR - http://www.scopus.com/inward/record.url?scp=85051642736&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85051642736&partnerID=8YFLogxK

U2 - 10.1016/j.bone.2018.04.020

DO - 10.1016/j.bone.2018.04.020

M3 - Article

C2 - 29698784

AN - SCOPUS:85051642736

VL - 116

SP - 207

EP - 214

JO - Bone

JF - Bone

SN - 8756-3282

ER -