Ateneo Lab Develops AI Dental Assistant for Accurate Diagnosis

Key Takeaways:

• The Ateneo Laboratory for Intelligent Visual Environments (ALIVE) has developed YOLO 11n, an AI software that detects odontogenic sinusitis with 98.2% accuracy.
• The technology uses dental X-rays and is designed to aid faster and cost-effective diagnosis, particularly in resource-limited areas.
• Collaboration between dentists and otolaryngologists is often required for traditional diagnosis, leading to delays in treatment.
• Researchers emphasize the continued need for human intervention alongside AI tools.
• The team aims for future commercialization but acknowledges regulatory hurdles before clinical use.

AI Technology Enhances Dental Diagnostics

The Ateneo Laboratory for Intelligent Visual Environments (ALIVE) has introduced a new AI-based tool, YOLO 11n, capable of detecting odontogenic sinusitis through dental X-rays with near-perfect accuracy, according to an article published on ABS-CBN News by Sam Beltran. This technological advancement, achieved in collaboration with researchers from Taiwan’s Chang Gung Memorial Hospital, National Cheng Kung University, Chung Yuan Christian University, and Ming Chi University of Technology, marks a significant step forward in addressing a condition often misdiagnosed due to its overlapping symptoms with general sinusitis.

Odontogenic sinusitis, caused by complications related to upper teeth, presents symptoms such as nasal congestion, foul-smelling nasal discharge, and occasional tooth pain. However, these signs are frequently overlooked by general practitioners, delaying appropriate treatment. Dr. Patricia Angela Abu, head of ALIVE, explained that traditional diagnosis requires collaboration between dentists and otolaryngologists, which can further prolong the process.

The YOLO 11n model offers a solution by swiftly analyzing X-ray images in real time, distinguishing affected areas with precision. This innovation not only accelerates diagnosis but also reduces the need for additional imaging techniques like CT scans, thereby lowering costs and minimizing radiation exposure for patients.

Training and Testing Phase: Challenges and Opportunities

Dr. Abu revealed to ABS-CBN News that training the AI model required fewer than 100 X-ray images—a relatively small dataset compared to other machine learning applications. However, obtaining these images posed challenges due to data privacy concerns and non-disclosure agreements. Despite this limitation, the model demonstrated remarkable accuracy during testing.

She emphasized that while AI holds promise in medical diagnostics, it is not intended to replace human expertise. “Yeah, it’s promising. It shows there are many biomedical imaging journals that make use of machine learning and deep learning to detect, recognize, categorize diseases, diagnosis, and many other things biomedical related,” Dr. Abu said.

The research team is currently focused on refining the technology and preparing it for broader application. They aim to eventually integrate YOLO 11n into clinical settings, although Dr. Abu acknowledged that regulatory standards must first be met. “It would be ideal if we can actually already have it used by medical doctors. But of course, it also needs to pass through certain metrics if a machine or a product or a tool is something that meets the standards and can be used already for clinical diagnosis,” she added.

Implications for Healthcare Accessibility

One of the most notable advantages of YOLO 11n is its potential to improve healthcare accessibility, particularly in underserved regions. By reducing the reliance on expensive diagnostic tools and specialized personnel, the AI assistant could enable quicker interventions and prevent complications that might otherwise spread to critical areas such as the face, eyes, or brain.

Additionally, the tool aligns with global efforts to leverage artificial intelligence in addressing public health challenges. As highlighted in the journal Bioengineering, where the findings were published, machine learning and deep learning applications continue to expand their role in disease detection and classification.

While the commercialization of YOLO 11n remains a work in progress, its development underscores the growing synergy between technology and healthcare. The Ateneo-led initiative serves as a reminder of how interdisciplinary collaboration can yield impactful innovations, even amidst limited resources and regulatory considerations.

This report is based on an article published by ABS-CBN News, authored by Sam Beltran.