File Download
  Links for fulltext
     (May Require Subscription)
Supplementary

Article: Deep learning predicts the malignant-transformation-free survival of oral potentially malignant disorders

TitleDeep learning predicts the malignant-transformation-free survival of oral potentially malignant disorders
Authors
KeywordsArtificial intelligence
Machine learning
Oral cancer
Oral leukoplakia
Oral lichenoid lesions
Issue Date2021
PublisherMDPI. The Journal's web site is located at http://www.mdpi.com/journal/cancers/
Citation
Cancers, 2021, v. 13 n. 23, p. 6054 How to Cite?
AbstractMachine-intelligence platforms for the prediction of the probability of malignant transformation of oral potentially malignant disorders are required as adjunctive decision-making platforms in contemporary clinical practice. This study utilized time-to-event learning models to predict malignant transformation in oral leukoplakia and oral lichenoid lesions. A total of 1098 patients with oral white lesions from two institutions were included in this study. In all, 26 features available from electronic health records were used to train four learning algorithms-Cox-Time, DeepHit, DeepSurv, random survival forest (RSF)-and one standard statistical method-Cox proportional hazards model. Discriminatory performance, calibration of survival estimates, and model stability were assessed using a concordance index (c-index), integrated Brier score (IBS), and standard deviation of the averaged c-index and IBS following training cross-validation. This study found that DeepSurv (c-index: 0.95, IBS: 0.04) and RSF (c-index: 0.91, IBS: 0.03) were the two outperforming models based on discrimination and calibration following internal validation. However, DeepSurv was more stable than RSF upon cross-validation. External validation confirmed the utility of DeepSurv for discrimination (c-index-0.82 vs. 0.73) and RSF for individual survival estimates (0.18 vs. 0.03). We deployed the DeepSurv model to encourage incipient application in clinical practice. Overall, time-to-event models are successful in predicting the malignant transformation of oral leukoplakia and oral lichenoid lesions.
DescriptionOpen Access Journal
Persistent Identifierhttp://hdl.handle.net/10722/310115
ISSN
2023 Impact Factor: 4.5
2023 SCImago Journal Rankings: 1.391
PubMed Central ID
ISI Accession Number ID

 

DC FieldValueLanguage
dc.contributor.authorAdeoye, JA-
dc.contributor.authorKoohi-Moghadam, M-
dc.contributor.authorLo, WIA-
dc.contributor.authorTsang, RKY-
dc.contributor.authorChow, LYV-
dc.contributor.authorZheng, L-
dc.contributor.authorChoi, SW-
dc.contributor.authorThomas, P-
dc.contributor.authorSu, Y-
dc.date.accessioned2022-01-24T02:24:02Z-
dc.date.available2022-01-24T02:24:02Z-
dc.date.issued2021-
dc.identifier.citationCancers, 2021, v. 13 n. 23, p. 6054-
dc.identifier.issn2072-6694-
dc.identifier.urihttp://hdl.handle.net/10722/310115-
dc.descriptionOpen Access Journal-
dc.description.abstractMachine-intelligence platforms for the prediction of the probability of malignant transformation of oral potentially malignant disorders are required as adjunctive decision-making platforms in contemporary clinical practice. This study utilized time-to-event learning models to predict malignant transformation in oral leukoplakia and oral lichenoid lesions. A total of 1098 patients with oral white lesions from two institutions were included in this study. In all, 26 features available from electronic health records were used to train four learning algorithms-Cox-Time, DeepHit, DeepSurv, random survival forest (RSF)-and one standard statistical method-Cox proportional hazards model. Discriminatory performance, calibration of survival estimates, and model stability were assessed using a concordance index (c-index), integrated Brier score (IBS), and standard deviation of the averaged c-index and IBS following training cross-validation. This study found that DeepSurv (c-index: 0.95, IBS: 0.04) and RSF (c-index: 0.91, IBS: 0.03) were the two outperforming models based on discrimination and calibration following internal validation. However, DeepSurv was more stable than RSF upon cross-validation. External validation confirmed the utility of DeepSurv for discrimination (c-index-0.82 vs. 0.73) and RSF for individual survival estimates (0.18 vs. 0.03). We deployed the DeepSurv model to encourage incipient application in clinical practice. Overall, time-to-event models are successful in predicting the malignant transformation of oral leukoplakia and oral lichenoid lesions.-
dc.languageeng-
dc.publisherMDPI. The Journal's web site is located at http://www.mdpi.com/journal/cancers/-
dc.relation.ispartofCancers-
dc.rightsCreative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License-
dc.subjectArtificial intelligence-
dc.subjectMachine learning-
dc.subjectOral cancer-
dc.subjectOral leukoplakia-
dc.subjectOral lichenoid lesions-
dc.titleDeep learning predicts the malignant-transformation-free survival of oral potentially malignant disorders-
dc.typeArticle-
dc.identifier.emailTsang, RKY: rkytsang@hku.hk-
dc.identifier.emailChow, LYV: chowlyv@HKUCC-COM.hku.hk-
dc.identifier.emailZheng, L: lwzheng@hkucc.hku.hk-
dc.identifier.emailChoi, SW: htswchoi@hku.hk-
dc.identifier.emailSu, Y: richsu@hku.hk-
dc.identifier.authorityTsang, RKY=rp01386-
dc.identifier.authorityZheng, L=rp01411-
dc.identifier.authorityChoi, SW=rp02552-
dc.identifier.authoritySu, Y=rp01916-
dc.description.naturelink_to_OA_fulltext-
dc.identifier.doi10.3390/cancers13236054-
dc.identifier.pmid34885164-
dc.identifier.pmcidPMC8657223-
dc.identifier.hkuros331442-
dc.identifier.volume13-
dc.identifier.issue23-
dc.identifier.spage6054-
dc.identifier.epage6054-
dc.identifier.isiWOS:000735226300001-
dc.publisher.placeSwitzerland-

Export via OAI-PMH Interface in XML Formats


OR


Export to Other Non-XML Formats