Comparison of the efficacy of periodontal prognostic systems in predicting tooth loss

Saydzai, Selai; Buontempo, Zoe; Patel, Pankti; Hasan, Fatemah; Sun, Chuanming; Akcalı, Aliye; Lin, Guo‐Hao; Donos, Nikos; Nibali, Luigi

doi:10.1111/jcpe.13672

Cited by 20 publications

(20 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Hence, in periodontics, it is very common to develop models with very poor sensitivity while also resulting in an apparently good general prognostic performance. In support of this concept, a recent study evaluated prospectively four different periodontal prognostic systems and found very high values of specificity but very low sensitivity F I G U R E 1 Area under the receiver operating characteristics curve for the different machine-learning models on the aggregate external validation cohort of molars predicting the outcome overall molar loss ranging between 3% and 12% (Saydzai et al, 2022). The challenge in developing accurate prediction models for the prognostic prediction in periodontology is to have a good balance between specificity and sensitivity.…”

Section: Discussionmentioning

confidence: 97%

“…The present study utilized these recommendations in an international collaboration with the aim of aggregating different cohorts and increasing the sample size for model development and validation (Rischke et al, 2022). These cohorts had previously been used for other prognostic studies published in periodontology (Shi et al, 2020; Petsos et al, 2021; Saleh et al, 2021; Saydzai et al, 2022). A higher rate of tooth loss appeared in the Beijing dataset, as well as some differences in predictor rates.…”

Section: Discussionmentioning

confidence: 99%

“…Future studies should compare the performance of artificial intelligence-based models with the traditional prognostic models. Indeed, although efforts have been made towards creating and comparing intelligence-based models (Bashir et al, 2022) as well as traditional prognostic methods (Saleh et al, 2021(Saleh et al, , 2022Saydzai et al, 2022), to the best of our knowledge no comparison between traditional versus intelligencebased models has been published.…”

Section: Discussionmentioning

confidence: 99%

See 2 more Smart Citations

Development and international validation of logistic regression and machine‐learning models for the prediction of 10‐year molar loss

Troiano

Nibali

Petsos

et al. 2022

J Clinic Periodontology

Self Cite

View full text Add to dashboard Cite

Aim: To develop and validate models based on logistic regression and artificial intelligence for prognostic prediction of molar survival in periodontally affected patients.Materials and Methods: Clinical and radiographic data from four different centres across four continents (two in Europe, one in the United States, and one in China) including 515 patients and 3157 molars were collected and used to train and test different types of machine-learning algorithms for their prognostic ability of molar loss over 10 years. The following models were trained: logistic regression, support vector machine, K-nearest neighbours, decision tree, random forest, artificial neural network, gradient boosting, and naive Bayes. In addition, different models were aggregated by means of the ensembled stacking method. The primary outcome of the study was related to the prediction of overall molar loss (MLO) in patients after active periodontal treatment. Results:The general performance in the external validation settings (aggregating three cohorts) revealed that the ensembled model, which combined neural network and logistic regression, showed the best performance among the different models for the prediction of MLO with an area under the curve (AUC) = 0.726. The neural network model showed the best AUC of 0.724 for the prediction of periodontitis-related molar loss. In addition, the ensembled model showed the best calibration performance.Conclusions: Through a multi-centre collaboration, both prognostic models for the prediction of molar loss were developed and externally validated. The ensembled model showed the best performance in terms of both discrimination and validation, and it is made freely available to clinicians for widespread use in clinical practice.

show abstract

Section: Discussionmentioning

confidence: 97%

Section: Discussionmentioning

confidence: 99%

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Development and international validation of logistic regression and machine‐learning models for the prediction of 10‐year molar loss

Troiano

Nibali

Petsos

et al. 2022

J Clinic Periodontology

Self Cite

View full text Add to dashboard Cite

show abstract

“…In the dental field, especially periodontology, prognostic systems have been developed based on teeth-related and periodontal predictors. However, to date, no highly sensitive prognostic system has been established [ 1 ]. Even with the support of machine learning methods, higher AUC and accuracy were not obtained compared with simpler methods [ 13 ].…”

Section: Discussionmentioning

confidence: 99%

“…Tooth loss can generally be prevented if disease is diagnosed and treated at an early stage for both caries and periodontal disease. However, while current periodontal disease prognostic systems are reported to have some success and reproducibility in predicting tooth loss and tooth retention, low sensitivity defines these models [ 1 ]. Thus, it is essential to develop better prognostication tools based on tooth mortality, particularly in the field of periodontology.…”

Section: Introductionmentioning

confidence: 99%

Machine Learning in Predicting Tooth Loss: A Systematic Review and Risk of Bias Assessment

Hasuike

Watanabe

Wakuda

et al. 2022

JPM

View full text Add to dashboard Cite

Predicting tooth loss is a persistent clinical challenge in the 21st century. While an emerging field in dentistry, computational solutions that employ machine learning are promising for enhancing clinical outcomes, including the chairside prognostication of tooth loss. We aimed to evaluate the risk of bias in prognostic prediction models of tooth loss that use machine learning. To do this, literature was searched in two electronic databases (MEDLINE via PubMed; Google Scholar) for studies that reported the accuracy or area under the curve (AUC) of prediction models. AUC measures the entire two-dimensional area underneath the entire receiver operating characteristic (ROC) curves. AUC provides an aggregate measure of performance across all possible classification thresholds. Although both development and validation were included in this review, studies that did not assess the accuracy or validation of boosting models (AdaBoosting, Gradient-boosting decision tree, XGBoost, LightGBM, CatBoost) were excluded. Five studies met criteria for inclusion and revealed high accuracy; however, models displayed a high risk of bias. Importantly, patient-level assessments combined with socioeconomic predictors performed better than clinical predictors alone. While there are current limitations, machine-learning-assisted models for tooth loss may enhance prognostication accuracy in combination with clinical and patient metadata in the future.

show abstract

Periodontal risk score: Initiation and model validation for 6,762 teeth

Saleh

Tattan

Troiano

et al. 2022

Journal of Periodontology

View full text Add to dashboard Cite

Background Tooth‐level prognostic systems are valuable tools for treatment planning and risk assessment of periodontally involved teeth. Recently the Miller‒McEntire prognosis index was found to outperform comparable systems. However, it had some limitations. The present study aimed to develop and evaluate the prognostic performance of a modified version that addresses most limitations of the previous model called the periodontal risk score (PRS). Methods Data were retrieved retrospectively from patients who received surgical and non‐surgical periodontal treatment at a university setting. Data on medical history and smoking status at baseline and the last maintenance visit were collected. Both univariate and multivariate Cox proportional hazard regression models were used to analyze the prognostic capability for predicting tooth loss due to periodontitis (TLP) risk. Results A total of 6762 teeth (281 patients) were followed up for a mean period of 22.6 ± 6.34 (10–47.6y) years. The PRS was successfully able to stratify the risk of TLP at baseline when the three different classes of association were compared for anterior and/or posterior tooth loss. After controlling for maintenance, age, and sex, the index showed an excellent predictive capacity for TLP with a Harrell C‐index of 0.947. Conclusions The periodontal risk score (PRS) displayed excellent predictive capability for anterior as well as posterior tooth loss due to periodontitis. This system was able to predict long‐term tooth loss with a very high accuracy in a population treated mainly by dental students and periodontics residents. The use of this/similar prognosis system is advisable as a means to establish tooth prognosis.

show abstract

Comparison of the efficacy of periodontal prognostic systems in predicting tooth loss

Cited by 20 publications

References 34 publications

Development and international validation of logistic regression and machine‐learning models for the prediction of 10‐year molar loss

Development and international validation of logistic regression and machine‐learning models for the prediction of 10‐year molar loss

Machine Learning in Predicting Tooth Loss: A Systematic Review and Risk of Bias Assessment

Periodontal risk score: Initiation and model validation for 6,762 teeth

Contact Info

Product

Resources

About