Gum disease generally encompasses a larger purview of periodontal disease which by its definition affects the supporting apparatus of the teeth. It affects a significant portion of the global population and is largely considered to be preventable. When left untreated, periodontal disease eventually leads to tooth loss with its consequential effects on both oral and systemic health. A number of systemic diseases and conditions have been observed to possess a relationship with periodontal disease, which include diabetes, cardiovascular disease, pregnancy, respiratory illnesses and Alzheimer’s disease to name a few. These disease processes are generally linked and mediated by the process of inflammation.1
Periodontal disease is insidious in its onset with its initial stages remaining largely asymptomatic. This, along with the known preventable nature of disease and its prevalence, makes timely detection, referral and intervention of paramount importance. Periodontal disease can pose a diagnostic challenge particularly in the background of varied and complex systemic interactions this disease process manifests. The complexity of its ramifications, the subjectivity of diagnostic parameters and concerns regarding intra- and inter-operator reliability beckon the involvement of artificial intelligence (AI) to standardize and simplify such issues. 2
Role of AI in gum disease
Diagnosis and treatment of periodontal disease involves the utilization of radiographic imaging as well as clinical methods such as periodontal probing depth measurement, detection of bleeding on probing and tooth mobility.
Introduction of artificial intelligence (AI) can aid in increasing the
subjectivity of such assessments along with potentially enhancing the
sensitivity and specificity of the diagnostic tools being implemented.
AI can also be for clinical demonstrations, treatment planning and mock-up surgeries supporting decision-making in periodontics. 3
Managing Referrals with Diagnocat AI
There can be instances in clinical settings in which general practitioners may have to decide referrals. This may prove to be challenging, particularly with the new periodontal disease classification system which involves a more detailed assessment to stage and grade the disease process.
An AI algorithm simplifies things by standardizing measurements and radiograph assessments to provide practitioners with an accurate diagnosis which is repeatable and objective, thereby essentially eliminating the somewhat steep learning curve required to make such evaluations.
There are also challenges in inter-operator and even intra-operator reliability of measurements such as probing depths and calculating clinical attachment levels. This is another sphere where AI can help eliminate the subjectivity of certain steps and provide quick, standardized solutions to generally time-taking clinical scenarios.
Diagnocat AI is based on SOTA computer vision deep learning models
which make it particularly adept at managing image-based workflows.
It offers a web-based interface which accepts a variety of radiographic configurations. Dentists can utilise this tool irrespective of the type of source and data being generated at the point of care. 4 Diagnocat AI has demonstrable efficacy in evaluating apical periodontitis and differentiating such lesions from healthy teeth. 3 It has also been evidenced to provide benefits in the detection of bone loss in periodontal lesions based on input data from panoramic radiographs.
Conclusion
AI seemingly appears to be the future in periodontal disease diagnosis. It has the potential, some of which has already been realized, to provide reliable, objective and repeatable diagnostic results. An accurate diagnosis is the first step in rendering the correct treatment to patients and by ensuring greater reliability in this regard, AI is set to eliminate a lot of doubt and subjective decision-making from the clinical set up, which can then enable clinicians to concentrate on delivering care to the patients.
Future Outlook
AI in periodontal disease is still a growing field, however, tools like Diagnocat AI appear to be a step in the right direction due to the sheer simplicity of their usage and minimal user-end standardization. Further research needs to look at the specificity and sensitivity of this tool for a variety of radiographic inputs in different lesions. Incorporating deep learning techniques can potentially train these algorithms to enhance their diagnostic efficacy. Further, clinical decision support tools based on SOTA-based methodologies can prove invaluable in primary care with potential in fortifying referral practices.
References
- Sahni V, Van Dyke TE. Immunomodulation of periodontitis with SPMs. Front Oral Health. 2023; 4:1288722. doi: 10.3389/froh.2023.1288722. PMID: 37927821; PMCID: PMC10623003.
- Scott J, Biancardi AM, Jones O, Andrew D. Artificial Intelligence in Periodontology: A Scoping Review. Dent J (Basel). 2023 Feb 8;11(2):43. doi: 10.3390/dj11020043. PMID: 36826188; PMCID: PMC9955396.
- Orhan K, Aktuna Belgin C, Manulis D, Golitsyna M, Bayrak S, Aksoy S, et al. Determining the reliability of diagnosis and treatment using artificial intelligence software with panoramic radiographs. Imaging Sci Dent. 2023; 53(3):199-208. doi: 10.5624/isd.20230109. Epub 2023 Aug 2. PMID: 37799743; PMCID: PMC10548159.
- Ezhov M, Gusarev M, Golitsyna M, Yates JM, Kushnerev E, Tamimi D, et al. Clinically applicable artificial intelligence system for dental diagnosis with CBCT. Sci Rep. 2021; 11(1):15006. doi: 10.1038/s41598-021-94093-9. Erratum in: Sci Rep. 2021 Nov 9;11(1):22217. doi: 10.1038/s41598-021-01678-5. PMID: 34294759; PMCID: PMC8298426.
