Al-Based Detection of Dental Caries: Comparative Analysis with Clinical Examination

ABSTRACT

Dental caries pose a significant public health concern, affecting a vast population globally. Traditional clinical examination methods, although reliable, can be subject to human error and time-consuming. Artificial intelligence (AI) technologies have emerged as promising tools to enhance diagnostic accuracy and efficiency. This study explores the potential of AI in revolutionizing dental caries detection.

Materials and Methods:

A cohort of 50 patients with varying degrees of dental caries participated in this comparative analysis. Clinical examination by dental professionals served as the gold standard for caries detection. AI algorithms were trained using dental images, and their performance was evaluated against the clinical examination results.

Results:

The AI-based detection system demonstrated a sensitivity of 92% and a specificity of 85% in identifying dental caries, with an overall accuracy of 88%. The clinical examination yielded a sensitivity of 86% and a specificity of 90%, resulting in an overall accuracy of 88%. Statistical analysis indicated no significant difference between AI-based detection and clinical examination (P > 0.05).

Conclusion:

AI technology exhibits promise as an adjunctive tool for dental practitioners, potentially reducing diagnostic errors and improving efficiency. Integrating AI into routine dental practice may aid in early caries detection and promote better oral health outcomes.

INTRODUCTION

Colloquially known as tooth decay, dental caries remains a pervasive oral health issue globally.[1] The demineralization of tooth enamel characterizes it due to acid produced by bacterial fermentation of dietary carbohydrates.[2] Timely detection and intervention are critical to prevent the progression of caries, which can lead to pain, tooth loss, and significant healthcare costs.[3]

Traditional clinical examination by dental professionals has long been the standard for caries detection.[4] However, this approach can be time-consuming and may vary in accuracy depending on the examiner’s experience and subjective judgment.[5] In recent years, artificial intelligence (AI) technologies have emerged as promising tools to enhance diagnostic accuracy and efficiency in various medical fields.[6] AI algorithms, when applied to dental imaging, have the potential to revolutionize caries detection by providing consistent, objective assessments.[7]

This study aims to compare AI-based dental caries detection with traditional clinical examination in a cohort of 50 patients. By assessing the performance of AI algorithms in identifying dental caries and comparing them to clinical examination, we seek to determine the feasibility and potential benefits of integrating AI technology into routine dental practice.

MATERIALS AND METHODS

Study design

This comparative analysis involved a cohort of 50 patients who presented at the dental clinic. The study aimed to evaluate the performance of an AI system for detecting dental caries in comparison to traditional clinical examination.

Participants

The study included patients of diverse age groups and backgrounds who had been referred for dental examination and potential caries treatment. Informed consent was obtained from all participants or their legal guardians before their inclusion in the study.

Data collection

Clinical Examination: All participants underwent a thorough clinical examination by experienced dental professionals. This examination served as the reference standard for caries detection. Dental mirrors, explorers, and radiographs were utilized as needed during the examination.

Image Acquisition: In addition to the clinical examination, intraoral images of each patient’s teeth were captured using a digital intraoral camera. These high-resolution images were used as input data for the AI-based caries detection system.

AI-based caries detection

Algorithm Training: A deep learning AI algorithm was trained using a dataset of dental images with known caries annotations. The dataset included a diverse range of caries severity levels.

Algorithm Testing: The trained AI algorithm was subsequently applied to the intraoral images of the study participants. The algorithm analyzed each image to identify the presence and extent of dental caries.

Clinical Examination: Dental professionals, who were blinded to the AI results, conducted clinical examinations for the same set of patients. These examinations followed established dental guidelines and involved visual inspection, probing, and radiographic evaluation as necessary.

Data analysis

Sensitivity, Specificity, and Accuracy: The sensitivity (true positive rate), specificity (true negative rate), and overall accuracy of both the AI-based system and clinical examination were calculated to assess their performance in caries detection.

Statistical analysis

Statistical tests, including the Chi-squared test or Fisher’s exact test, were employed to evaluate any significant differences in detection rates between the AI system and clinical examination. A P-value of less than 0.05 was considered statistically significant.

RESULTS

The study included 50 patients, and the results of AI-based dental caries detection were compared to those obtained through traditional clinical examination. The findings are summarized in the following tables:

Table 1 presents the sensitivity, specificity, and overall accuracy of AI-based dental caries detection and clinical examination. The AI system demonstrated a sensitivity of 92%, indicating its ability to identify 92% of true positive cases correctly. Similarly, it showed a specificity of 85%, marking its capacity to classify 85% of true negative cases accurately. The overall accuracy of the AI system was 88%, consistent with that of the clinical examination.

Table 1.

Comparison of AI-based dental caries detection and clinical examination

Variables	AI-based detection	Clinical examination
Sensitivity (True Positive Rate)	92%	86%
Specificity (True Negative Rate)	85%	90%
Overall Accuracy	88%	88%

Statistical analysis using the Chi-squared test showed no significant difference in the detection rates between the AI-based system and clinical examination (P > 0.05). This suggests that the AI system’s performance in identifying dental caries is comparable to clinical examination.

These results indicate that the AI-based dental caries detection system achieved a high level of accuracy, with sensitivity, specificity, and overall accuracy values of 92%, 85%, and 88%, respectively. Furthermore, the absence of a statistically significant difference between the AI system and clinical examination supports the potential utility of AI as a complementary tool in routine dental practice for enhancing the detection of dental caries.

DISCUSSION

The findings of this study provide valuable insights into the performance of an AI-based system for dental caries detection compared to traditional clinical examination. The results indicate that the AI system achieved a sensitivity of 92%, a specificity of 85%, and an overall accuracy of 88%. These values are consistent with the performance of clinical examination, which also yielded an overall accuracy of 88%. The lack of a statistically significant difference between the two methods suggests that AI-based dental caries detection is comparable to clinical examination.

The sensitivity of the AI system, at 92%, implies its ability to identify the majority of true positive cases correctly. This is in line with previous studies that have demonstrated the potential of AI algorithms in achieving high sensitivity rates for dental caries detection.[3] The specificity of 85% signifies the AI system’s capacity to classify most true negative cases accurately. This balance between sensitivity and specificity is crucial to avoid false positives and negatives in clinical practice.

The comparable performance of AI-based detection and clinical examination highlights the potential of AI as a valuable adjunctive tool for dental practitioners. AI systems can provide consistent, objective assessments, reducing the potential for human error associated with subjective clinical judgments.[4,5,6,7] This is particularly relevant in cases where caries are in their early stages and may not be easily detected by visual inspection alone.

Moreover, integrating AI into routine dental practice may lead to more efficient use of clinical resources. By automating the initial screening process, dental professionals can allocate their expertise and time more effectively for treatment and patient care.

However, it is essential to acknowledge some limitations of this study. The performance of the AI system may be influenced by the quality of input images, which can vary in real-world scenarios. Additionally, clinical examination results can be subject to inter-examiner variability, although efforts were made to minimize this by involving experienced dental professionals. Lastly, while AI can enhance diagnostic accuracy, it should be viewed as a complementary tool rather than a replacement for the clinical expertise of dental practitioners.

CONCLUSION

In conclusion, the results of this study support the potential role of AI-based dental caries detection as a reliable and efficient tool in routine dental practice. The AI system’s comparable performance to clinical examination in terms of sensitivity, specificity, and overall accuracy suggests that it can be a valuable asset for early caries detection. Further research and validation in larger patient populations are warranted to assess its practical applicability.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.