Machine learning is a method of data analytics that involves teaching computers to do what humans naturally do, which is learning from experience. Algorithms developed in machine learning involve the use of computational techniques so that the computer can directly learn information using data without reliance upon a preset equation. As the amount of data increases, algorithms are able to progressively improve their performance.
Tooth decay and periodontal disease are known factors in predicting whether a given individual may experience tooth loss in the future. The loss of teeth can have a profound effect on a patient who may experience a loss of self-esteem and a decline in their well-being.
The Harvard School of Dental Medicine published the results of a recent study that they conducted on June 18, 2021, in PLOS ONE. The researchers developed a set of five algorithms based upon their study of almost 12,000 individuals. Some factors that went into the development of the different algorithms included the socioeconomic status of the studied individual, their education level, their income level, race, and the presence of arthritis and diabetes.
The machine learning techniques developed in the study were hoped to provide the ability to assess a given individual’s susceptibility to experiencing tooth loss. Those high-risk individuals could then receive a subsequent physical examination and early diagnosis to provide them with their best chance of retaining their natural teeth.
Hawazin Elani, the lead investigator, indicated that machine-learning techniques can be used to predict risk, but those incorporating variables like socioeconomic status are particularly useful in screening and identifying individuals with a greater risk of experiencing tooth loss. In fact, it is hoped that the results of the study can even help non-professionals to identify at-risk patients.
The study itself provided data that allowed researchers to design the machine-learning algorithms. Researchers then assessed the ability of the different algorithms to predict both incremental and complete tooth loss in different individuals.
Remarkably, the machine-learning algorithms provided the ability to assess tooth loss risk without a physical examination. However, those identified as being in the high-risk group would be referred for an actual physical exam.
According to Elani, the algorithm models that incorporated socioeconomic traits offered improved prediction capability over those that only relied upon regular clinical indicators. So taking a patient’s education level, income, and employment status, into account is just as important as relying upon an assessment of their clinical dental status.
Since early detection of tooth loss risk can lead to an earlier diagnosis and treatment, it is believed that the machine-learning algorithms developed in the Harvard study will lead to improved outcomes for those at greatest risk of experiencing tooth loss.
At Kitchener Dentist Centre, we care about your oral health and strive to provide you with the best possible treatment and care. If you have any questions or concerns, please contact us.
DISCLAIMER: The advice offered is intended to be informational only and generic in nature. It is in no way offering a definitive diagnosis or specific treatment recommendations for your particular situation. Any advice offered is no substitute for proper evaluation and care by a qualified dentist.
