Using AI to improve decision-making in periodontal care

PerioPredict began as part of a broader customer relationship management (CRM) platform for dental clinics, with the goal of providing each user with tools tailored to their clinical needs. As the periodontology module of this platform, PerioPredict was built to support structured decision-making in periodontal care.

What started out as a focused feature rapidly evolved into a standalone innovation, which offers periodontists a new way to assess periodontal health, classify disease, and forecast treatment outcomes with greater clarity and efficiency.

At its core, PerioPredict enables clinicians to input structured clinical data—including measurements, photographs, and contextual notes—and receive real-time insights. The system is AI-assisted rather than AI-powered. This distinction is important: the platform relies on manual input to ensure accuracy and to maintain clinician oversight, while machine-learning tools are used to accelerate data entry and enhance analysis. This hybrid model ensures the clinician remains in control, supported by technology that simplifies and strengthens their process.

PerioPredict offers two main capabilities:

1. It helps categorize the patient’s current periodontal status using structured clinical criteria, improving consistency in classification (using the staging and grading model from the 2018 classification) and speed in documentation.
2. It uses machine-learning models trained on real-world clinical cases to provide predictive insights—such as the likelihood of successful gingival-recession coverage or the stability of clinical attachment over time.

One of our long-term ambitions for the platform is to introduce a new classification model altogether, one that levers multiple dimensions of clinical data—far beyond the limits of traditional frameworks—to offer a more nuanced picture of periodontal health, risk, and predictive outcomes.

A significant challenge has been standardizing the input of periodontal data, particularly when it involves combining clinical images with charted metrics. While building an intuitive interface, we also had to ensure that the platform aligns with real-world clinical workflows and meets evolving expectations around data security and regulatory compliance.

Recent developments in AI infrastructure—including federated learning and lightweight edge models—will allow us to expand while preserving privacy, security, and accuracy. We are now working on a redesigned architecture that will support clinics globally, with robust controls for anonymization and data ownership.

Winning the EFP Digital Innovation Award has provided us with crucial validation and has opened doors to broader collaboration and strategic development. This recognition has come at a pivotal time. Rather than simply continuing the current version, we are using this momentum to scale PerioPredict into a more advanced, flexible, and research-grade platform.

The ultimate goal is to make PerioPredict an integral part of modern periodontal care—not just as a digital assistant, but as a catalyst for better, evidence-based treatment planning. For researchers and clinicians alike, it offers a way to bridge daily practice with broader clinical insights. For potential partners and investors, it represents a scalable solution grounded in clinical need, with the flexibility to grow alongside rapid technological change.

As we move into the next phase of development, we are actively exploring collaborations, strategic partnerships, and investment opportunities to help scale PerioPredict into a global platform for periodontal innovation.