AI-based medical diagnostics company AIDOT has registered a key patent in Vietnam for its AI-powered automated cervical cancer diagnosis system, establishing a strong intellectual property (IP) defense framework in the Southeast Asian market.
AIDOT recently announced that it received patent registration (Patent No. VN 52703) from the National Office of Intellectual Property of Vietnam (NOIP) for its “Automatic Cervical Cancer Diagnosis System.”
The patent, initially filed in 2019, was granted in September 2025 and will be protected for 20 years until 2039.
Unlike a simple AI algorithm, this patent protects the “entire system.” The most notable feature of the registered patent is that it is a system-level patent, covering the full structure of a commercial medical system rather than a single AI algorithm component.
The patent defines the entire workflow of AI diagnosis in actual clinical environments, including:
- Generation of learning data from cervical images
- Preprocessing of captured images
- Machine learning-based automatic diagnosis
- User interface (UI) for medical staff and users
- Retraining incorporating specialist evaluation
- Storage of diagnostic and evaluation information
AIDOT’s Research Director, Choi Han-sol, explained, “This patent does not merely claim the concept or algorithm of diagnosing cancer with AI, but rather covers the structure itself as used in real medical settings.” He emphasized, “Even small modifications would be difficult to bypass, and infringement determination will be clear.”
In other words, this patent is designed so that competitors cannot easily circumvent it even if they alter some functions or the UI.
The independent claims broadly define the core structure of the automatic diagnostic system, while the dependent claims specify elements used in actual clinical practice, such as:
- Color and shape-based classification
- Acetic acid (VIA) and Lugol’s iodine reactions
- Retraining structure based on specialist feedback
Accordingly, if competitors in Vietnam adopt a system that uses a VIA (visual inspection) + AI combination, or classify and diagnose cervical images with AI, display results on a screen, and incorporate them into learning, the likelihood of clear patent infringement determination is high.
An intellectual property industry expert commented, “This patent cannot be avoided simply by changing some algorithms. Implementing the same diagnostic workflow would constitute infringement, effectively blocking copy products in the Vietnamese market.”
In short, the Cerviray AI system’s features—image-based automatic diagnosis, AI classification, specialist input, and continuous learning—are directly reflected in the patent claims, providing a solid legal foundation for commercialization in Vietnam.
AIDOT is already closely collaborating with the Vietnamese Ministry of Health (MOH), university hospitals, and public screening projects. The patent registration is expected to significantly enhance credibility for national pilot programs and Official Development Assistance (ODA) projects.
Through this Vietnam patent, AIDOT aims to block domestic copy products and low-cost Chinese AI solutions, securing exclusive status in public screenings and Vietnamese operations. The company noted that the patent clearly addresses the common “copy and imitation risk” in the medical AI field, effectively blocking it at the system level, which has been highly evaluated by global investors.
AIDOT CEO Jeong Jae-hoon stated, “The Vietnam patent is not just about securing rights—it is a core infrastructure for public screening and global business expansion. We intend to dominate the market through genuine innovation rather than copying.”
Korea Economic TV / December 16, 2025 / Reporter Yang Jae-jun
