Although the association between risk factors and non-surgical root canal treatment (NSRCT) failure has been extensively studied, methods to predict the outcomes of NSRCT are in an early stage, and dentists currently make the treatment prognosis based mainly on their clinical experience. Since this involves different sources of error, we investigated the use of machine learning (ML) models as a second opinion to support the clinical decision on whether to perform NSRCT. We undertook a retrospective study of 119 confirmed and not previously treated Apical Periodontitis cases that received the same treatment by the same specialist. For each patient, we recorded the variables from a newly proposed data collection template and defined a binary outcome: Success if the lesion clears and failure otherwise. We conducted tests for detecting the association between the variables and the outcome and selected a set of variables as the initial inputs into four ML algorithms: Logistic Regression (LR), Random Forest (RF), Naive-Bayes (NB), and K Nearest Neighbors (KNN). According to our results, RF and KNN significantly improve (p-values < 0.05) the sensitivity and accuracy of the dentist’s treatment prognosis. Taking our results as a proof of concept, we conclude that future randomized clinical trials are worth designing to test the clinical utility of ML models as a second opinion for NSRCT prognosis.

Background/Objectives: In a previous study, we utilized categorical variables and machine learning (ML) algorithms to predict the success of non-surgical root canal treatments (NSRCTs) in apical periodontitis (AP), classifying the outcome as either success (healed) or failure (not healed). Given the importance of radiographic imaging in diagnosis, the present study evaluates the efficacy of deep learning (DL) in predicting NSRCT outcomes using two-dimensional (2D) periapical radiographs, comparing its performance with ML models. Methods: The DL model was trained and validated using leave-one-out cross-validation (LOOCV). Its output was incorporated into the set of categorical variables, and the ML study was reproduced using backward stepwise selection (BSS). The chi-square test was applied to assess the association between this new variable and NSRCT outcomes. Finally, after identifying the best-performing method from the ML study reproduction, statistical comparisons were conducted between this method, clinical professionals, and the image-based model using Fisher’s exact test. Results: The association study yielded a p-value of 0.000000127, highlighting the predictive capability of 2D radiographs. After incorporating the DL-based predictive variable, the ML algorithm that demonstrated the best performance was logistic regression (LR), differing from the previous study, where random forest (RF) was the top performer. When comparing the deep learning–logistic regression (DL-LR) model with the clinician’s prognosis (DP), DL-LR showed superior performance with a statistically significant difference (p-value < 0.05) in sensitivity, NPV, and accuracy. The same trend was observed in the DL vs. DP comparison. However, no statistically significant differences were found in the comparisons of RF vs. DL-LR, RF vs. DL, or DL vs. DL-LR. Conclusions: The findings of this study suggest that image-based artificial intelligence models exhibit superior predictive capability compared with those relying exclusively on categorical data. Moreover, they outperform clinician prognosis.

Background/Objectives: Managing caries lesions that affect the inner third of the dentin is crucial to ensuring pulp vitality; the clinician must make decisions that will affect the vitality of the tooth. Our purpose is to understand the behavior of Spanish dentists in treating deep cavities and to examine whether variations exist based on their academic training and/or years of professional experience. Methods: This study was approved by the ethics committee of the Balearic Islands CEI-IB. A survey was conducted using the SurveyMonkey platform with 11 questions, the first 4 of which focused on defining the characteristics of the respondents. The following six concerned a clinical case of deep caries in tooth number 4.7, and the last regarded the opinion of the actual treatment of the case. The survey was sent by email in April 2022. The results were analyzed with the SPSS 29.0 program using the chi-square test. Results: A total of 347 responses were obtained (93.95%), and those surveyed stated that they apply minimal intervention concepts in their treatments, with 90.49% performing conservative dentistry treatments daily. A total of 56.48% of the respondents had bachelor’s degrees, 12.39% had graduated, 33.14% had a postgraduate degree, 38.90% had a master’s degree, and 17% had a doctorate. Most (40.63%) had been in professional practice for 16–30 years. Conclusions: Significant differences were identified regarding years of professional experience in terms of decision-making in methods of treatment and the choice of materials used for pulp protection. Likewise, significant differences were found regarding the academic training of the respondents, the cavity cleaning method selected, and the use of chemical substances for removing carious dentin. We can conclude that academic training and years of professional practice influence decision-making at some points in treating deep caries lesions.

Background/Objectives In a previous study, we utilized categorical variables and Machine Learning (ML) algo-rithms to predict the success of non-surgical root canal treatments (NSRCTs) in apical periodontitis (AP), classifying the outcome as either success (healed) or failure (not healed). Given the importance of radiographic imaging in diagnosis, the present study evaluates the efficacy of Deep Learning (DL) in predicting NSRCT outcomes using two-dimensional (2D) periapical radiographs, comparing its performance with ML models. Methods The DL model was trained and validated using Leave-One-Out Cross-Validation (LOOCV). Its output was incorporated into the set of categorical variables, and the ML study was reproduced using Backward Stepwise Selection (BSS). The chi-square test was applied to assess the association between this new variable and NSRCT outcomes. Finally, after identifying the best-performing method from the ML study reproduction, statistical comparisons were conducted between this method, clinical professionals, and the im-age-based model using Fisher’s Exact Test. Results The association study yielded a p-value of 0.000000127, highlighting the predictive ca-pability of 2D radiographs. After incorporating the DL-based predictive variable, the ML algorithm that demonstrated the best performance was Logistic Regression (LR), differing from the previous study, where Random Forest (RF) was the top performer. When comparing the Deep Learning – Logistic Regression (DL-LR) model with the clinician’s prognosis (DP), DL-LR showed superior performance with a statistically significant dif-ference (p-value < 0.05) in sensitivity, NPV, and accuracy. The same trend was observed in the DL vs. DP comparison. However, no statistically significant differences were found in the comparisons of RF vs. DL-LR, RF vs. DL, and DL vs. DL-LR. Conclusions The findings of this study suggest that image-based artificial intelligence models exhibit superior predictive capability compared to those relying exclusively on categorical data. Moreover, they outperform clinician prognosis.