Maintaining proper dental hygiene is vital for overall health, with early detection of dental diseases playing a key role in preventing serious health issues. Although X-rays are the current gold standard for diagnosing dental conditions, their accessibility remains limited for many individuals worldwide.
In a collaborative effort, researchers from Carnegie Mellon College of Engineering and the University of Pittsburgh School of Dental Medicine have developed an innovative dental health sensing system. This system utilizes commercially available electric toothbrushes for at-home dental condition assessments, with their findings published in the journal Proceedings of the ACM on Interactive, Mobile, Wearable and Ubiquitous Technologies.
According to the Centers for Disease Control, around 57 million Americans reside in areas lacking adequate dental health professionals, with approximately 67% of these shortages occurring in rural regions. The aim of this initiative is to empower individuals with an electric toothbrush capable of performing self-examinations, ultimately providing essential dental care to millions who might otherwise go without.
Dubbed the ToMoBrush (Tooth Monitoring Brush / Tomorrow’s Toothbrush), this innovative tool repurposes a standard electric toothbrush, requiring only minimal hardware modifications, to facilitate regular at-home dental assessments.
Rather than viewing a toothbrush solely as a cleaning tool, the ToMoBrush exploits the acoustic signals produced by the rapid vibrations of its bristles. When the brush makes contact with a tooth, the tooth vibrates in sync, generating distinct acoustic signals that reflect each tooth’s condition.
“Dental disease poses a significant public health challenge, leading to pain, infections, and complications with eating, speaking, and social interaction,” says Kuang Yuan, a Ph.D. student in electrical and computer engineering. “We aimed to develop a cost-effective solution for dental health monitoring that users can implement regularly in the comfort of their homes.”
The research team established a data-driven signal processing pipeline to identify and differentiate various dental conditions, including cavities, plaque buildup, and food impaction. The system also accounts for variations among electric toothbrushes, such as brand differences, battery charge, and bristle design. To manage these variables, the researchers modeled the vibration dynamics involving the toothbrush, tooth resonance, brushing force, and movement.
In their paper, the researchers propose an algorithm designed to isolate these factors and extract clear tooth resonance signatures based on a crucial observation: although these factors occupy the same frequency band, their rates of change across frequencies differ. By employing techniques commonly used in speech processing to differentiate between glottal excitation and vocal tract resonances, the team transforms the signal into the cepstrum domain, where these unique characteristics become more easily distinguishable.
“After obtaining the tooth resonance signature, we developed a feature selection algorithm to identify specific regions of the signature that are particularly effective in detecting three distinct dental conditions,” Yuan explains. “We can conduct health assessments by comparing the signatures with previously established healthy reference measurements.”
The research team envisions that such a system could enhance the dental healthcare landscape, providing early alerts for potential issues between professional visits, even for those who have access to dental services.
Related topics:
